On the Capacity of Energy Harvesting Communication Link Mehdi Ashraphijuo, Vaneet Aggarwal, Senior Member, IEEE, and Xiaodong Wang, Fellow, IEEE

Size: px
Start display at page:

Download "On the Capacity of Energy Harvesting Communication Link Mehdi Ashraphijuo, Vaneet Aggarwal, Senior Member, IEEE, and Xiaodong Wang, Fellow, IEEE"

Transcription

1 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER On the Caacity of Energy Harvesting Communication Lin Mehdi Ashrahijuo, Vaneet Aggarwal, Senior Member, IEEE, and Xiaodong Wang, Fellow, IEEE Abstract We consider an energy harvesting oint-to-oint communication system where the transmitter is owered by an energy arrival rocess and is equied with a battery of finite caacity B max, which could be used for saving energy for future use. We assume a discrete i.i.d. energy arrival rocess where at each time ste, energy of amount A i is harvested with robability i i {,,...,K} indeendent of the other time stes. We rovide uer and lower bounds on the caacity of this channel. These bounds are shown to be within a constant ga for K 3for all arameters, and for K > 3 when the battery caacity B max is small or large enough, where this constant does not deend on any energy or battery arameters. Index Terms Energy harvesting, channel caacity, achievability, uer bound, energy arrival rocess. I. INTRODUCTION W IRELESS communication networs that are comosed of devices that can harvest energy from nature, reresent the green future of wireless systems. The simlest model that catures the communication scenario is a discrete-time AWGN channel. In most communication systems, the transmission ower is a major cost [] [4]. This cost can be artially alleviated by using energy harvesting devices [5] [8]. For transmitters that are owered by energy harvesters, energy is tyically relenished by the energy harvester and exended for communications or other rocesses; any unused energy is then stored in an energy storage, such as a rechargeable battery [9], [0]. The i.i.d. arrival rocess of the incoming energy in energy harvesting is a good starting oint for general results and is a common assumtion in literature [5] [7], [0], []. However, unlie conventional communications where devices are only subject to a ower constraint or a total energy constraint, transmitters with energy harvesting caabilities are subject to additional energy harvesting constraints [] [4]. It is noteworthy to mention other directions in the literature on characterizing the energy harvesting communication system caacity such as Manuscrit received March 8, 05; revised June 4, 05; acceted Setember 3, 05. Date of ublication Setember 5, 05; date of current version November 6, 05. The material in this aer was resented in art at the Annual Allerton Conference on Communication, Control, and Comuting, Chamaign, IL, USA, October 05. This wor was suorted in art by the U.S. National Science Foundation under grant CCF-565. M. Ashrahijuo is with the Electrical Engineering Deartment, Columbia University, New Yor, NY 007 USA mehdi@ee.columbia.edu. V. Aggarwal is with the Industrial Engineering Deartment, Purdue University, West Lafayette, IN USA vaneet@urdue.edu. X. Wang is with the Electrical Engineering Deartment, Columbia University, New Yor, NY 007 USA wangx@ee.columbia.edu. Color versions of one or more of the figures in this aer are available online at htt://ieeexlore.ieee.org. Digital Object Identifier 0.09/JSAC acet scheduling otimality [5], [6] and throughut analysis with unreliable energy sources [7], [8]. Secifically, in every time ste, the transmitter is constrained to use at most the amount of stored energy that is currently available, although more energy may become available in the future slots. Thus, a causality constraint is imosed on the use of the harvested energy. This raises many interesting issues in the design of efficient energy harvesting communication schemes. In this aer, we study the bounds on the caacity of AWGN channel where the transmit node is owered by stochastic energy arrivals, and is equied with the caability of storing and drawing energy from a finite caacity battery. The caacity of a single-user channel with infinite battery caacity is obtained in [9], where it is shown that the caacity equals to that of a classical AWGN channel with an average ower constraint equal to the average energy harvesting rate. Desite the recent significant efforts to characterize the caacity of the energy harvesting channel in the finite battery case [0] [], there is still a lac of comlete understanding. For examle, [] rovides a formulation of the caacity and derives a lower bound on the caacity that is only numerically comutable. However, it is difficult to obtain useful insights from numerical evaluations. Even in the case of no battery, where [3] rovides an exact single-letter characterization of the caacity in terms of an otimization roblem, the corresonding otimization is difficult to solve and requires numerical evaluations. The authors of [4] investigated the case of constant inut energy with a limited battery. Recently, the aroximate caacity of a single-user energy harvesting system with discrete energy arrivals has been characterized in [8], where it is assumed that all the incoming energy is stored in the battery and the transmission energy is taen from the battery. We note that [8] considers a similar energy harvesting model with the difference that the incoming energy cannot be used directly when it arrives at the transmitter and consequently if there is not enough sace in the battery it will be wasted. In contrast, we assume that the incoming energy can be used for transmission and consequently increases the flexibility of the transmitter, which rovides the ossibility of achieving better rates. The results in [8] have been further extended to fading channels in [5], [6]. We assume that the energy arrival is an i.i.d. discrete random rocess that taes K values of A,...,A K, where energy A i arrives with robability i. The incoming energy in art is used for transmission and in art is stored into the battery, which has a caacity B max. We rovide uer and lower bounds on the caacity of such channel when the receiver does not IEEE. Personal use is ermitted, but reublication/redistribution requires IEEE ermission. See htt:// for more information.

2 67 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 now the energy arrival rocess. These bounds are within a constant ga for K 3.04 bits for K,.884 bits for K and 4.46 bits for K 3, where the constant does not deend on any energy or battery arameters. For K > 3, the ga between the bounds is shown to be a constant that is indeendent of all arameters in the cases of small enough B max A A and large enough B max A K K i i A i battery sizes. Our aroximate caacity characterizations rovide imortant insights on the otimal design of energy harvesting communication systems. For lower bound, we introduce multile strategies and choose the best for each set of system model arameters which maes our results different from that in [8] where the lower bound is based on a single strategy. For each strategy, we derive a unique energy allocation olicy that is time invariant and the consumtion of each arriving energy acage is decreasing over time with a geometric arameter across different eochs. The roosed uer bound accounts for the flexibility of incoming harvested energy, and thus needs to decide how much incoming energy to store in battery, and how much to utilize for transmission in each time instant which maes our results more comlex than that in [8] where all the incoming energy can only be stored in the battery. The remainder of this aer is organized as follows. Section II introduces the model for a oint-to-oint communication system, equied with an energy harvesting device. Sections III and IV give the results on the uer and lower bounds of the caacity, resectively. These bounds are shown to be within a constant ga in several cases in Section V. Finally, Section VI concludes the aer. II. SYSTEM MODEL We consider a oint-to-oint channel with a single transmitter, equied with an energy harvesting device that has a battery with a caacity of B max. If the battery is not full, the harvested energy is in art stored in the battery and in art used for transmission; if the battery is full, the transmitter can directly use all the harvested energy. In both cases, some additional amount of energy that is stored in the battery can also be used for transmission. Let X t denote the scalar real inut to the channel at time t. We consider a discrete-time AWGN channel, where the outut of the channel is given by Y t X t N t, where N t N0, is the additive white Gaussian noise. At each time t, the system harvests E t units of energy that is causally nown at the transmitter i.e., at time t the transmitter nows E t, E t,...butis not nown at the receiver. Let B t be the available energy in the battery at time t. We assume that at each time, the system first harvests energy and then transmits the signal X t. The square of X t is constrained by the available energy B t lus the harvested energy E t, i.e., X t B t E t. And the available battery energy B t is udated as B t min{b max, B t E t X t }. Note that in order to not waste the harvested energy, we should choose X t such that B t E t X t B max, which leads to the lower bound on X t, i.e., B t E t B max X t. 3 Here, we consider the case that the harvested energy E t is a K -level i.i.d. rocess as E t A with robability,,...,k, 4 where 0 A <...< A K, K and,..., K > 0. Definition : The encoding functions f t, t,, n and the decoding function g are defined as f t : M E t X, t,, n, 5 g : Y n M, 6 where X Y R, E {A,, A K } and M {,, M} is the set of messages to be transmitted. To transmit message w M, attimet,, n, the transmitter sends X t f t w, {E i } t i0. The battery state B t is a deterministic function of {X i } t i0, {E i} t i0, therefore also of w, {E i} t i0.the functions f t must satisfy the energy constraints and 3: B t w, {Ei } t i0 Et B max f t w, {Ei } t i0 Bt w, {Ei } t i0 Et. 7 The receiver estimates ŵ g {Y i } n i0. The robability of error is P n e M M P ŵ w w was transmitted. 8 w The rate of an M, n code is M n. We say rate R is achievable if for every δ>0 there exists, for all sufficiently large n, anm, n code with rate M n > R δ, and error P e n 0. The caacity C of the above system with arameters A,...,A K,,..., K and B max is defined as the suremum of all achievable rates. Remar : For the secial case of B max, the caacity of the above system has been characterized in [9]. It is shown that in the otimal transmission scheme, nothing is transmitted for the first few time slots so that enough energy is accumulated. This is followed by transmission using the average harvested energy in every time ste. Hence, the caacity is characterized by the average energy arrival rate. [3] also characterized the caacity with infinite buffer size using a different aroach. Remar : In [8], for the Bernoulli energy arrival, i.e., K, A 0, the aroximate caacity is characterized under a different battery model where the transmission energy can only be taen from the battery and thus even if A > B max,theextra energy, A B max, cannot be utilized. Remar 3: In the remainder of the aer, we assume K > since for K i.e., constant inut energy, A is an uer bound on the caacity since A is the average ower. Further, a lower bound on the caacity is A.04,

3 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 673 which can be achieved by the amlitude-constrained AWGN channel [8, Lemma ], leading to a maximal ga of.04 bits between the uer and lower bounds. The authors of [4] investigated the case of constant inut energy with a finite-caacity battery and obtained better bounds on caacity, where the lower bound and uer bound are quite close. III. UPPER BOUND An uer bound on the caacity of the discrete-time AWGN channel with an energy harvesting transmitter is given as follows. Theorem : For K, 0 A <...<A K, K,,..., K > 0 and B max 0, the caacity is uer bounded by C ub,k which is the solution to the following otimization roblem: C ub,k max S K z,...,z K z,...,z K subject to z i min{b max, A i }, i,...,k, K i z i 0, 9 i where S K z,...,z K K i i A i z i Ki A i z i. Proof: The claimed uer bound holds intuitively, because at times of energy arrival A i, i >, if the average energy that is ut in the battery is z i, then the caacity in these time slots is uer bounded by A i z i. The energy z i taen out from the incoming energy and stored in the battery can be utilized in the time slots with energy arrival of A. And consequently, an average ower constraint forms the uer bound. Define gt Xt as the ower allocation strategy that maximizes the long-term average transmission rate over the class of feasible online olicies and and also define g i t gt Et A i. Then, the caacity is uer bounded as C lim inf N E a lim inf N E [ N N t [ K N i [ b lim inf E K N c K i i gt N t ] ] g it ] N i N N g i t, N i i t E [ g i t ], 0 where N i is the number of occurrences of E t A i for t {,...,N}, and as N, N i N i. In the above, a follows by searating the incoming energies by their levels, b follows from the concavity of the function and c follows from the law of large numbers. Suose that an average of x i energy is stored into the battery and y i i energy is drawn from the battery when the energy arrival is A i for i >, then Eg i t i A i x i y i for i > and Eg t A K i i x i y i.asn,usingthe law of large numbers, it can be concluded that the caacity is uer bounded by C max 0x i,y i,x i y i B max, 0 K i i x i y i { K i i A i x i y i } Ki i x i y i A. By defining z i x i y i we get the result stated in the theorem. The following two results follow from Theorem. Corollary : For the case of K, the uer bound can be reduced to { C ub, max 0xB max A x A x }. The otimal value of x in is x min{b max, A A }. The otimum solution to 9 for general K is given exlicitely by the following theorem. The roof is given in Aendix A. Theorem : The exlicit uer bound is given as follows for different ranges of B max. A For B max A A, K l C ub,k A l B max l A B Kj max j. 3 s B For A s r r s i i A i B max A s sr r si i A i, s K, K l Kis C ub,k A i l B max ls si i A i B Kjs max j. 4 K ms m C For A K K i i A i B max, C ub,k K i A i. 5 i IV. THE LOWER BOUND The insights develoed in derivation of the uer bound can be used to give an achievability scheme, which achieves the rate given in the following theorem.

4 674 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 Fig.. Illustration of the energy allocation olicy for C 3 lb,4 x with q q 3 and r q 3. Note that the energy control olicy is reset each time a acet A 3 or A 4 arrives. Theorem 3: For K, 0 A <...<A K, K,,..., K > 0, and B max 0 define q K i i. Then, therateof max max,...,k 0x min{b max,a } C lb,k x can be achieved, where K Clb,K x j A j x j h A h xq q.884 K K >. 6 The rest of this section roves this theorem. We show the achievability of Clb,K x, for any {,...,K } and x [0, min{b max, A }]. Let the time slots of two consecutive energy arrivals of at least A be T and T, i.e., E T, E T A and E t < A for T < t < T. Then, the energy of E T x is used at time T and the energy of x goes into the battery. At any time t between T and T, the energy of q B t is extracted from the battery, and thus the energy of E t q B t can be used for transmission, where E t < A, and the residual energy at time t in the battery is B t q B t. Thus, the energy usage from the battery at time t {T,...,T } is q q tt x. We note that this is a geometric random variable with arameter q, and thus has the mean q. We ignore the battery energy residue at T, and consider the next interval starting at T to find the energy usage after T. This olicy can be evaluated to give the desired bound. The energy utilization strategy g t roosed above is of the form g t φ j, where j t max{τ : E τ A i, i, τ t}, i.e., the strategy is invariant across time and the allocated energy deends on the number of time stes since the last energy arrival A i, i. The random variable φ j is defined as q q j x A, w.. q, φ j., q q j x A, w.. q, for all j, 7 and A x, w.. q, φ A K x, w.. K q, Fig. deicts an examle of energy allocation olicy for Clb,4 3 x. The idea for the achievability scheme is that if both the transmitter and receiver now at each time arrival what energy acet A j arrives, they can agree on an energy allocation strategy ahead of time. Communication roceeds as follows: At each time ste t,the transmitter sees the realization of the energy rocess E t,let j t max{t t : E t A }, i.e., the number of time stes since the last time battery was recharged with energy acet arrival A j, j. Letφ i denote the amount of energy allocated to transmission, i channel uses after the last time the battery was recharged via acet arrivals A j, j, i 0,,. We concentrate on an energy allocation olicy φ i that is invariant across different eochs the eriod of time between two adjacent acet arrivals A j and A j, j, j. In other words, if energy A j, j arrives at the current channel use, we allocate φ 0 amount of energy for transmission; if energy A j, j arrived in the revious channel use but not the current channel use, then we allocate φ amount of energy for transmission, and so on till the next arrival of energy A j, j. Consider n consecutive time slots of energy arrivals where n is large enough. Denote n i, i 0 as the number of times slots t that their corresonding j t max{t t : E t A } is equal to i.ifn goes to infinity it can be shown that n i, i 0 goes to infinity, as well. The transmitter and the receiver agree on a sequence of M codeboos: C 0, C, C,, CM with large enough M, each codeboo C i consisting of ni R i codewords where R i is the rate of the codeboo and codeboo C i is amlitude-constrained to φ i, i.e., the symbols of each codeword in C i are such that X t φ i if i t max{t t : E t A }. This ensures that the symbol transmitted at the corresonding time will not exceed the energy

5 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 675 constraint φ i. The transmitter chooses a codeword c,i C i, i {0,,, M} to communicate to the receiver. More secifically, in the l th occurrence of i, the transmitter sends the l th symbol of codeword c,i, i.e., uon the arrival of the first energy acet A j, j, the transmitter sends the first symbol of c,0 C 0 ; if there is no energy acet arrival A j, j in the next channel use, it transmits the first symbol of c, C in the next channel use, etc. Once the second energy acet A j, j arrives, the transmitter transmits the second symbol of c,0, then the second symbol of c,,etc.if j > M, the transmitter transmits zero symbol. Communication ends when the transmitter observes the arrival of the n 0 th energy acet of energy A j, j. We assume that communication starts with the arrival of the first energy acet of energy A j, j. We assume that HE t bits is used to communicate the incoming energy level E t to the receiver. The receiver can trac the codeboo used by the transmitter and decode each codeword searately by nowing the energy arrival E t in the transmitter. Let {S l} l L be the inter-arrival times between the lth and l th energy arrivals of E t A, where L is the total number of energy arrivals of E t A between t and t N, i.e., L l S l N < L l S l. Notice that S l s are i.i.d. geometric random variables with arameter q, and thus has the mean q. We can lower bound the rate achieved by g t in terms of these new variables as lim inf N N N t lim inf L a E b q E g t L S l l j0 E[ φ j] L l S l [ S [ j0 E φ j i i E[S ] i PS i j0 j0 ]] E φ j i q q q i E φ j j0 i j q q i E q q j E j0 φ j φ j, 9 where a follows from the law of large numbers and b follows from E[S ] q. Using Eqn. 9 and the discussed energy allocation strategy before it, the rate in Lemma below could be achieved, which will be further lower bounded to get the bound in the statement of the theorem. First, we give the following lemma which will be used in the roof of Lemma. Lemma : [8, Lemma ] Lower bound on amlitudeconstrained AWGN caacity max I X; Y A x:x A Lemma : The caacity C of a system with i.i.d. energy arrival rocess that is only causally nown at the transmitter but not at the receiver is lower bounded by C q q j E φ j j0.04 H,..., K. Proof: We showed that 9 is achievable for an AWGN channel with channel state information i.e., energy arrival rocess at the receiver. For the model in this aer, the ga of.04 is due to amlitude-constrained AWGN given in Lemma, and H,..., K is due to having no channel state information at the receiver since the energy level can be communicated to the receiver in H,..., K bits. The next result further lower bounds the achievable rate in Lemma. The roof is given in Aendix B. Lemma 3: The following inequality holds for all 0 x min{b max, A }: RHS of K j h j A j x q A h q.884 K K >. Using Lemma and Lemma 3, Theorem 3 then follows. V. THE GAP BETWEEN THE BOUNDS In this section, we show that the ga between the uer and lower bounds is bounded by some constant in several cases, where the constant does not deend on any energy or battery arameters. A. Case of K From Theorems and 3, the following corollary follows. Corollary : For K, the uer and lower bounds on the caacity are within a constant ga. More formally, C ub, Clb, x.884 bits 3 for,, A, A and B max 0. Proof: From Theorem 3, we get the lower bound A x.884, 4 C lb, x A x for all 0 x min{b max, A }. The uer bound is given by Corollary, i.e., Eq. that together with 4 gives 3.

6 676 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 Fig.. Comarison of the bounds of the roosed model with those in [8] for K. We can further consider a more general case where the usage of the battery is not free. In this model, the cost of storing energy into the battery is denoted through the coefficient 0 < r in. Similarly, the cost of taing energy out the battery is denoted through the coefficient 0 < r out. Then Eqns. and become X t r out B t E t and B t B t r in E t X t r out X t E t, resectively. And the uer bound in Corollary and the constant ga result in Corollary can be generalized as follows. The roof is given in Aendix C. Theorem 4: For the case of K with the battery efficiency arameters r in and r out, we have the following uer bound on the caacity C ub, max 0xB max { A x A r out r in x }. 5 { where x min B max, A A r outr in r out r in }.Wealso have C ub,.884 C C ub,, 6 For K, assuming, A 0 and A 0 5, in Fig. we lot the uer and lower bounds for r in r out and r in r out 0.8, resectively, as well as the bounds in [8] for varying values of B max. The bounds decrease with a decrease in efficiencies r in and r out. Recall that [8] assumes a different system model from the one used in this aer, where the harvested energy cannot be directly used without being stored in the battery first. Since the achievability scheme in [8] is also a feasible strategy for our model, the ga can be imroved by choosing the maximum of the achievable rate in this aer and that in [8]. Fig. 3. Comarison of uer and lower bounds of the roosed model for K 3. B. Case of K 3 The next result gives the ga between the lower and uer bounds for K 3. Theorem 5: For K 3, the ga between the uer and the lower bounds is bounded by 4.46 bits. More formally } C ub,3 max {Clb,3 x, C3 lb,3 x bits 7 for,, 3, A, A, A 3 and B max 0. Proof: We note that for K 3 there is a constant of.884 K K > 3.96 in 6, and we can show that the ga between the uer bound C ub,3 given in Theorem and the maximum of the lower bounds Clb,3 x and Clb,3 3 x 3 given in 6 is at most 0.5 lus the above 3.96 bits. In order to show this ga of 0.5 bit, we consider 5 cases deending on the values of B max, A, A and A 3. The detailed ga evaluation for these five cases is given in Aendix D. For K 3, assuming 3 3, A 0, A 0 3 and A 3 0 6, in Fig. 3 we lot the uer and lower bounds as a function of B max. It is seen that both bounds increase with the battery caacity; the uer bound saturates at 3 i i A i 9.74 bit/s and the lower bound saturates at bit/s. C. Case of K > 3 The roosed achievability scheme can achieve a rate that is within a constant ga to the uer bound of the caacity for general K for arts A and C of Theorem as shown in the following theorem. Theorem 6: The following constant ga results for general K hold: For B max A A,wehave C ub,k C lb,k B max.34 K. 8

7 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 677 For A K K i i A i B max,wehave u C ub,k Clb,K u A u j.34 3 K, 9 j where u arg max i A i Kji j. Proof: Part : Follows from 3 and 6 that C ub,k C lb,k B max.34 K. 30 Part : Using 5 we have C ub,k K i A i i K a i A i KA u A u K j ji K j ju K j ju K, 3 where a follows from the fact that u arg max i A Kji i j. Moreover, using 6 we have u A u C u lb,k u j lu j j j A u Kmu u n n m A j u }{{} Au u Kmu n n m u K l u A l A u j.34 K j }{{} A u A u A u u j j K j ju.34 K. 3 For K 5, assuming, 5 8, A 0, A i 0.5i0 5, i {,...,5}, in Fig. 4, we lot the uer and lower bounds for a range of values of B max.forb max A A 5 0 4, an uer bound is given in Case A of Theorem and its ga to the achievable rate is shown to be within bits in Part of Theorem 6. Also, for B max A K K i i A i , an uer bound is H A z Kl A l z l 3 Kl A l z l. K Kl A l z l 3 A 3 A 3 z 3 Kl l z l 3 Kl A l z l. K 3 A Kl l z l K Kl A l z l 3 K Kl A l z l K A K z K K Kl A l z l a Kl A l z l 3 Kl A l z l. K Kl A l z l 3 Kl A l z l 3 Kl A l z l. K 3 Kl A l z l K Kl A l z l 3 K Kl A l z l K Kl A l z l A Kl [,..., K ] T [,..., K ] 0, 33 l z l

8 678 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 We also define μ gi,μ hi, i {,...,K },μ u as the corresonding Lagrangian multiliers of these constraints, resectively. A solution z,...,z K is otimum if it has the following roerties: K S K z,...,z K μ gi g i z,...,z K K i i μ hi h i z,...,z K μ u uz,...,z K, 37 μ u 0,μ gi 0,μ hi 0, for all i,...,k, μ gi g i z,...,z K 0, for all i,...,k, 38 μ hi h i z,...,z K 0, for all i,...,k, μ u uz,...,z K Fig. 4. Comarison of uer and lower bounds of the roosed model for K 5. given in Case C of Theorem and its ga to the achievable rate is shown to be within bits in Part of Theorem 6. In addition, for A A < B max < A K Ki i A i, an uer bound is given in Case B of Theorem and Clb,5 3 min{a j 3 j, B max } is considered as the lower bound. Further, it can be seen that both bounds increase with the battery caacity, and the uer bound saturates at 5 i i A i VI. CONCLUSIONS We have considered an energy-harvesting communication system where a transmitter owered by an exogenous energy arrival rocess, modeled as a discrete random rocess, and equied with a battery of finite caacity, communicates over a discrete-time AWGN channel. We have develoed uer and lower bounds on the caacity of such systems, which are shown to be within a constant ga for K 3, and some cases of K > 3. Extension to fading channels is a future wor. APPENDIX A PROOF OF THEOREM We first show that S K z,...,z K is concave by showing its Hessian matrix is negative-semidefinite. The Hessian is given by 33, as shown at the bottom of the revious age, where a follows from the fact that i < 0, for all i,...,k. A i z i Define the following functions corresonding to the constraints g i z,...,z K z i B max 0, i,...,k, 34 h i z,...,z K z i A i 0, i,...,k, 35 uz,...,z K z... K z K Now we rove the theorem for the given 3 cases: A For B max A A we show the otimality of z l B max, l {,...,K } which by substituting in the uer bound 9 results in 3. We have uz,...,z K B max < 0 and also for all i,...,k h i z,...,z K B max A i A A A i a < 0, 40 where a follows from the fact that A < A < A 3 <...< A K. So, we should have μ hi 0, i {,...,K },μ u 0. Also, g i z,...,z K 0, i {,...,K }. To show the otimality, it is sufficient to show that μ gi 0, for all i,...,k. Using these and 37, we get μ gi i A B max, A i B max for all i,...,k. Moreover, we have μ gi 0, for all i,...,k, since μ gi i A B A max i B max a A i A B max A B max A i B max A A B max A B max A i B max b 0, where a follows from A < A 3 <...<A K and b follows from B max A A. s B For A s r r s i i A i < B max A s sr r si i A i we show the otimality of sm zl A l m A m K ns n B max sl, l {,...,s} and l zm B max, l {s,...,k } which by substituting in

9 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 679 the uer bound 9 results in 4. For all i,...,s we have g i z,...,z K z i B max sm m A m K A i ns n B max sl l B max a < 0, where a follows from A s A s s r r r r s i i A i <...< s i i A i < B max.wealsohave h i z,...,z K z i A i sm m A m K ns n B max sl < 0, 4 l for all i,...,s. It can be also seen that for all i s,...,k h i z,...,z K z i A i B max A i s s A s r i A i A s r i K s i A i < 0. 4 In addition, we have A s K uz,...,z K j z j j r rs B max K s K i js j <0. 43 So, we should have μ g j 0, j {,...,s},μ hi 0, i {,...,K}, and μ u 0. Also, g i z,...,z K 0, i {s,...,k }. It is sufficient to show that μ gi 0, for all i s,...,k. Using these and 37, we get that the first s elements of S K z,...,z are 0 and also μ gi i sm m A m K ns n B max sl l, A i B max for all i s,...,k. Moreover, we have μ gi 0, for all i s,...,k, since 44, shown at the bottom of the age, holds where a follows from A s < A s <...< A K and b follows from B max A sr s r si i A i. C For A K K i i A i B max we show the otimality of zl A l K i i A i, l {,...,K } which by substituting in the uer bound 9 results 5. For this set of zi s, we get S K z,...,z K 0 which shows the otimality. APPENDIX B PROOF OF LEMMA 3 We divide the roof into two arts; K > and K. A. Proof for K > For the case of K >, we divide the roof into three cases. The first case is when q x q A > A, the second one is when q x q A A, and the third one is when q x q A q A q x q A q,for< q < and A.4. Let choice of A is a result of erforming a minimization of a function that will be mentioned in Remar 4. I For the case q x q A q A q x q A q,for< q <. We have K j j A j x q A h q K j hq h j A j x q j A j q A h q, 45 μ gi i sm m A m K ns n B max sl l A i B max sm m A m K ns n B max sl l A i B max sm m A m K ns n B max sl A i B max l A i s l l s m m A m B max sm s l l m A m K ns n B max sl A i B max l a A s s l l s m m A m B max b sm s l l m A m 0, 44 K ns n B max sl A i B max l

10 680 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 and j0 q q j E φ j a q h K n n q A nx q q q j j } {{ } j K n q q j j q q j x A h }{{} > q q jx Ah q q jx Ah q j q j q x q A h hq jq q j q x A h q hq q 46 n A nx, 47 where a follows from 7 and 8. Using 45 and 46 we get K j j A j x q A h q h q q j E φ j q j j0 hq q j j j a q j 0.7 j A hq q x A h q jq q j hq j A hq jq q j hq j A hq q A h q q q q x q A h q x q A h q j q j b q j q j A ln hq j A ln hq j.4 ln hq q x q A h A j , 49 where a follows since j jq q j q c q d q q 0.7, and b follows from A.4. To show c, we use the identity j0 jq q j q E[X] q q, where X geometricq, and j0 q q j and d follows from the fact that Gq q q q is a continuous bounded function of q 0,. Furthermore, it is monotonically decreasing and is uer bounded by lim q 0 Gq ln 0.7. Remar 4: The choice of A.4 is made since it minimizes the RHS of 48. In other words, for q { { }} j j, arg min A max 0 A A ln.4, and thus choosing A.4 results in the best bound using 48. II For the case q x q A >.4, we have: j0 q q j E φ j K j A j x q A h q j h a q q x A h A h q q h }{{} jq q j q j q x h q A h jq q j q j q j q x q A q x q A h jq q j q 50

11 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 68 b q q x 0.7 q A q q ln x 0.7 q A , 5 ln.4 where a follows from 46 and b follows from Ste a used in 48. III For the case q x q A.4 we have: K j j A j x q A h q j0 a h h q q j E φ j q A h q q q j j h q q j x A h h h q A h q , 53 where a follows from 46. So, we get K j j A j x q A h q j0 h q q j E φ j.30, 54 which together with H,, K K comletes the roof for K >, i.e., by comaring this to the statement of the lemma and the fact that.884 K H,..., K.30 K H,..., K.30, the result is roven. when x A A, where A.4. The choice of constant.4 is exlained in Remar 5. I For the case x A > A we have: K j j A j x h j0 q A h q q q j E φ j H, a q q x q A q q q x A H, ln x A H, H, ln A H, ln A H, ln A b.5.844, 55 ln A B. Proof for K For the case of K, we divide the roof into two cases. The first case is when x A > A and the second one is where a follows from 50, b is due to the fact that the exression is a continuous bounded function of 0,. Furthermore, it is concave and attains a maximum value of ln A.5 at 0.43.

12 68 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 II For the case x A A we have: K j A j x j q A h q h q q j E φ j h j0 a b q A h H, q x A H, H, A H A.844, 56 where a follows from 5 and b follows from the fact that K. Remar 5: If the ga result in the two cases is et as function of A rather than substituting the value as in the roof above, ln A.5 from 55 is a strictly decreasing function of A and A from 56 is a strictly increasing function of A. Since the ga is the maximum of two gas, otimizing the maximum of the gas gives A.4 using which we get both the RHS of 55 and the RHS of to be equal. So, for K we get K j j A j x q A h q j0 h q q j E φ j H,.844, 57 which together with H comletes the roof for K, i.e., by comaring this to the statement of the lemma and the fact that , the result follows. APPENDIX C PROOF OF THEOREM 4 We divide the roof of Theorem 4 into two arts, corresonding to the uer and lower bounds, resectively. C. Uer Bound Similar to the roof of Theorem, suose that an average of x energy is stored into the battery and y energy is drawn from the battery when the energy arrival is A, then Eg t A x r out y and Eg t A r out r in x r in y.asn, using the law of large numbers, we see that the caacity is uer bounded by { C max 0x B max, A x r out y 0 r out r in x r y in A r out r in x r } in y. 58 Now, we relace the variable x by x x min{x, y r out r in } and also the variable y by y y r out min{x, y r out r in }. We see that A x r outy A r out r in x y r in A x r out y A r out r in x y r in r out r in min{x, y r out r in } r out r in A x r out y A r out r in x y. 59 r in Thus, C ub, with x and y is less than or equal to C ub, with x and y if min{x, y r out r in } > 0. So, for otimal x and y,min{x, y r out r in }0. Since r out r in x r in y 0, we get y 0 and thus the uer bound in 5 follows. D. Lower Bound To show the achievability given in 6, we show that the rate of max x can be achieved, where 0xmin{B max,a } C lb, Clb, x A x A r inr out x Let the time slots of two consecutive energy arrivals of A be T and T, i.e., E T, E T A and E t A for T < t < T. Then, the energy of E T x is used at time T and the energy of r in x goes into the battery is. At any time t between T and T, the energy of B t is extracted from the battery, and thus the energy of E t r out B t can be used for transmission, where E t A, and the residual energy at time t in the battery is B t B t. Thus, the energy usage from the battery at time t {T,...,T } is r in r out tt x. We note that this is a geometric random variable with arameter, and thus has the mean. We ignore the battery energy residue at T, and consider the next interval starting at T to find the energy usage after T.

13 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 683 The energy utilization strategy gt roosed above is of the form gt φj, where j t max{τ : E τ A, τ t}, i.e., the strategy is invariant across time and the allocated energy deends on the number of time stes since the last energy arrival of A. The random variable φj is defined as φj r in r out j x A, for all j, and φ0 A x. 6 The rest of the roof is straightforward and in the same lines as in the case of r in r out as in Section IV, with the only difference that in the roof of Lemma 3 in Aendix VI the choice of two cases change as follows. The first case is when r in r out x A >.4 and the second one is when r in r out x A.4. and Clb,3 3 A 3 A 3 A 3 A 3 A A 3 A 3 A A 3 A 3 A 3.96 A 3 A 3 A So, the maximum of the rates in 64 and 65 is achievable because of B max A 3 A A A and it is seen from 63 that their maximum is within a constant ga of 4.46 bits to the uer bound. APPENDIX D PROOF OF THEOREM 5 Case : B max A A : The uer bound in 3 can be achieved within 3.96 bits using the achievability scheme Clb,3 B max in Theorem 3. Case : B max A 3 A : We have Then using 5 we get A A A 3 3 A A 3 A 3 3 A A 3 A A 3 A 3 A max{a A 3 A,A 3 A 3 A }. 6 C ub,3 A A A 3 3 max{a A 3 A, A 3 A 3 A }. 63 Also, we have the following two achievable rates Clb,3 A A 3 A 3 A A A A A A 3 A A 3.96 A A 3 A 3.96, 64 3 Case 3: A A <B max < A 3 A and A A 3 A A 3 A 3 A : For this case, similar to the revious case, we get C ub,3 A A A 3 3 max{a A 3 A, A 3 A 3 A } A A 3 A, 66 and the rate in 64 is achievable because of B max A A and from 66 the maximum is within a constant ga of 4.46 bits to the uer bound. 4 Case 4: A A <B max < A 3 3 i i A i and A A 3 A <A 3 A 3 A : In this case for the uer bound in Theorem, we have z3 B max. So, C ub,3 max 0x B max { A x 3 A 3 B max A x 3 B max 3 A 3 B max A A A 3 B max }. 67 Now, divide the achievability scheme into two arts: A A 3 B max : The scheme Clb,3 3 x with x min{b max, A 3 }B max gives the caacity within 4.46 bits to the uer bound 67 because Clb,3 3 B max equates shown at the bottom of the next age. B A A > 3 B max : For this set of of arameters, we have > 3. This is because if 3,itresults

14 684 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 A A A A 3 > B max which is a contradiction to the initial assumtion of A A B max. Use the strategy Clb,3 3 A A for the achievability. Then, we get C ub,3 Clb,3 3 A A 3.96 A A {}}{ A A A 3 B max A A A A A A A A A 3 A A A A A A 3 68 a. 69 where a follows since 3 which can be easily shown. 5 Case 5: 3 i i A i < B max < A 3 A and A A 3 A <A 3 A 3 A : Tae the uer bound C ub,3 A A 3 A 3. For achievability consider Clb,3 3 A 3 3 i i A i. We could see that 3 x 3 i i A i A A, so we will have 3 A 3 i A i 3.96 C 3 lb,3 i 3 3 i A i i 3 i A i i 3 i A i i A A A A, and we get C ub Clb,3 3 A 3 c 3 i A i 3.96 i A A 3 i A i i A A 3 i A i a b lim A lim A max 3 3 e d { { { i A A 3 A 3 A A 3 A 3 A 70 A 3 A 3 } A A 3 A 3 A A 3 A 3 } A 7 A 3 A 3 3 e } 3 e 3 3 7, 73 where a follows from the fact that the function gx x x is increasing and also A A A 3 i i A i A 3 A 3 and b follows from the fact that 7 has a ositive derivative resect to A, c follows from 3 3 A 3 A and d follows from the fact that 7 has a negative derivative resect to e and e 3 3 in 73. Clb,3 3 B max A 3 B max A 3 B max A 3 B max. }{{ } A 3 B max A 3 B max A A A 3 B max

15 ASHRAPHIJUO et al.: ON THE CAPACITY OF ENERGY HARVESTING COMMUNICATION LINK 685 It remains to show that 73 /. We divide the roof into two arts: A : Let tae m n, m and 3 m n. Then, we get m mn 3. 3 m mnn 3 So, we have 3 3. B > :Lettae m, m n and 3 m n. Then, we get m m n nm n m m nm n nm n m nm n m nm n m m n max n { m m nm n m n m nm n }. 74 We see that otimal n, n m m m. Then, RHS of 74 is equal to m mm mmmm m, where 0 < m mm. It can be seen that m is an increasing function with resect mm to m, and is thus. ACKNOWLEDGMENT We would lie to than the anonymous reviewers for their comments and suggestions which significantly imroved the aer. REFERENCES [] S. Lindsey and C. Raghavendra, PEGASIS: Power-efficient gathering in sensor information systems, in Proc. IEEE Aeros. Conf., 00, vol. 3, [] W. Heinzelman, A. Chandraasan, and H. Balarishnan, Energyefficient communication rotocol for wireless microsensor networs, in Proc. 33rd Annu. Hawaii Int. Conf. Syst. Sci., Jan. 000,. 0. [3] R. Ramachandran, V. Sharma, and P. Viswanath, Caacity of Gaussian channels with energy harvesting and rocessing cost, IEEE Trans. Inf. Theory, vol. 60, no. 5, , Mar. 04. [4] X. Kang, Y.-K. Chia, C. K. Ho, and S. Sun, Cost minimization for fading channels with energy harvesting and conventional energy, in Proc. IEEE Int. Conf. Commun. ICC, 04, [5] D. Niyato, E. Hossain, M. Rashid, and V. Bhargava, Wireless sensor networs with energy harvesting technoies: A game-theoretic aroach to otimal energy management, IEEE Wireless Commun., vol. 4, no. 4, , Aug [6] A. Kansal and M. Srivastava, An environmental energy harvesting framewor for sensor networs, in Proc. Int. Sym. Low Power Electron. Des. ISLPED, Aug. 003, [7] M. Rahimi, H. Shah, G. Suhatme, J. Heideman, and D. Estrin, Studying the feasibility of energy harvesting in a mobile sensor networ, in Proc. IEEE Int. Conf. Robot. Autom. ICRA, 003, vol., [8] S. Sudevalayam and P. Kularni, Energy harvesting sensor nodes: Survey and imlications, IEEE Commun. Surveys Tuts., vol. 3, no. 3, , Third Quarter 0. [9] R. Rao, S. Vrudhula, and D. Rahmatov, Battery modeling for energy aware system design, Comuter, vol. 36, no., , 003. [0] D. Brunelli, L. Benini, C. Moser, and L. Thiele, An efficient solar energy harvester for wireless sensor nodes, in Proc. Des. Autom. Test Euroe, Mar. 008, [] W.-G. Seah, Z. Eu, and H. Tan, Wireless sensor networs owered by ambient energy harvesting WSN-HEAP Survey and challenges, in Proc. st Int. Conf. Wireless Commun. Veh. Technol. Inf. Theory Aeros. Electron. Syst. Technol. VITAE, 009,. 5. [] S. Chen, P. Sinha, N. Shroff, and C. Joo, Finite-horizon energy allocation and routing scheme in rechargeable sensor networs, in Proc. IEEE INFOCOM, Ar. 0, [3] Z. Wang, V. Aggarwal, and X. Wang, Iterative dynamic water-filling for fading multile-access channels with energy harvesting, IEEE J. Sel. Areas Commun., vol. 33, no. 3, , Mar. 05. [4] Z. Wang, V. Aggarwal, and X. Wang, Power allocation for energy harvesting transmitter with causal information, IEEE Trans. Commun., vol. 6, no., , Nov. 04. [5] J. Yang and S. Uluus, Otimal acet scheduling in an energy harvesting communication system, IEEE Trans. Commun., vol. 60, no.,. 0 30, Jan. 0. [6] M. A. Anteli, E. Uysal-Biyioglu, and H. Eral, Otimal acet scheduling on an energy harvesting broadcast lin, IEEE J. Sel. Areas Commun., vol. 9, no. 8,. 7 73, Se. 0. [7] C. K. Ho and R. Zhang, Otimal energy allocation for wireless communications with energy harvesting constraints, IEEE Trans. Signal Process., vol. 60, no. 9, , Se. 0. [8] Y. Dong, F. Farnia, and A. Ozgur, Near otimal energy control and aroximate caacity of energy harvesting communication, IEEE J. Sel. Areas Commun., vol. 33, no. 3, , Mar. 05. [9] O. Ozel and S. Uluus, Achieving AWGN caacity under stochastic energy harvesting, IEEE Trans. Inf. Theory, vol. 58, no. 0, , Oct. 0. [0] K. Tutuncuoglu, O. Ozel, A. Yener, and S. Uluus, Binary energy harvesting channel with finite energy storage, in Proc. IEEE Int. Sym. Inf. Theory ISIT, Jul. 03, [] P. K. Deeshith, V. Sharma, and R. Rajesh, AWGN channel caacity of energy harvesting transmitters with a finite energy buffer, arxiv rerint arxiv: , 03. [] W. Mao and B. Hassibi, On the caacity of a communication system with energy harvesting and a limited battery, in Proc. IEEE Int. Sym. Inf. Theory ISIT, Jul. 03, [3] O. Ozel and S. Uluus, AWGN channel under time-varying amlitude constraints with causal information at the transmitter, in Proc. Conf. Rec. 45 Asilomar Conf. Signals Syst. Comut. ASILOMAR, Nov. 0, [4] V. Jog and V. Anantharam, An energy harvesting AWGN channel with a finite battery, in Proc. IEEE Int. Sym. Inf. Theory ISIT, Jun. 04, [5] J. Doshi and R. Vaze, Long term throughut and aroximate caacity of transmitter-receiver energy harvesting channel with fading, in Proc. IEEE Int. Conf. Commun. Syst. ICCS, Nov. 04, [6] R. Nagda, S. Satathi, and R. Vaze, Otimal offline and cometitive online strategies for transmitter-receiver energy harvesting, arxiv rerint arxiv: 40.9, Oct. 04.

16 686 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO., DECEMBER 05 Mehdi Ashrahijuo received the B.Sc. and M.Sc. degrees in electrical engineering from Sharif University of Technoy, Tehran, Iran, in 00 and 0, resectively. He is currently ursuing the Ph.D. degree in electrical engineering from Columbia University, New Yor, NY, USA. His research interests include networ information theory and statistical signal rocessing with alications in wireless communication, digital communications, underwater sensor networs, and alications of the queuing theory in wireless sensor networs, develoing fundamental rinciles for communication networ design, with emhasis on develoing interference management and understanding the role of feedbac and cooeration. He was the reciient of the Qualcomm Innovation Fellowshi in 04. Vaneet Aggarwal S 08 M SM 5 received the B.Tech. degree from the Indian Institute of Technoy, Kanur, India, and the M.A. and Ph.D. degrees from the Princeton University, Princeton, NJ, USA, all in electrical engineering, in 005, 007, and 00, resectively. He is currently an Assistant Professor with the Purdue University, West Lafayette, IN, USA. Prior to this, he was a Senior Member of Technical Staff Research with AT&T Labs-Research, Middletown, NJ, USA, and an Adjunct Assistant Professor with the Columbia University, New Yor, NY, USA. His research interests include alications of information and coding theory to wireless systems and distributed storage systems. He was the reciient of the Princeton University s Porter Ogden Jacobus Honorific Fellowshi in 009. Xiaodong Wang S 98 M 98 SM 04 F 08 received the Ph.D. degree in electrical engineering from the Princeton University, Princeton, NJ, USA. He is a Professor of electrical engineering with Columbia University, New Yor, NY, USA. Among his ublications is a boo entitled Wireless Communication Systems: Advanced Techniques for Signal Recetion Prentice Hall in 003. His research interests include comuting, signal rocessing and communications, and he has ublished extensively in these areas, and wireless communications, statistical signal rocessing, and genomic signal rocessing. He has served as an Associate Editor for the IEEE TRANSACTIONS ON COMMUNICATIONS, the IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, the IEEE TRANSACTIONS ON SIGNAL PROCESSING, and the IEEE TRANSACTIONS ON INFORMATION THEORY. He was the reciient of the 999 NSF CAREER Award, the 00 IEEE Communications Society and Information Theory Society Joint Paer Award, and the 0 IEEE Communication Society Award for Outstanding Paer on New Communication Toics. He is listed as an ISI Highly-cited Author.

Improved Capacity Bounds for the Binary Energy Harvesting Channel

Improved Capacity Bounds for the Binary Energy Harvesting Channel Imroved Caacity Bounds for the Binary Energy Harvesting Channel Kaya Tutuncuoglu 1, Omur Ozel 2, Aylin Yener 1, and Sennur Ulukus 2 1 Deartment of Electrical Engineering, The Pennsylvania State University,

More information

Capacity of the Discrete Memoryless Energy Harvesting Channel with Side Information

Capacity of the Discrete Memoryless Energy Harvesting Channel with Side Information 204 IEEE International Symposium on Information Theory Capacity of the Discrete Memoryless Energy Harvesting Channel with Side Information Omur Ozel, Kaya Tutuncuoglu 2, Sennur Ulukus, and Aylin Yener

More information

Online Scheduling for Energy Harvesting Broadcast Channels with Finite Battery

Online Scheduling for Energy Harvesting Broadcast Channels with Finite Battery Online Scheduling for Energy Harvesting Broadcast Channels with Finite Battery Abdulrahman Baknina Sennur Ulukus Department of Electrical and Computer Engineering University of Maryland, College Park,

More information

ECE 534 Information Theory - Midterm 2

ECE 534 Information Theory - Midterm 2 ECE 534 Information Theory - Midterm Nov.4, 009. 3:30-4:45 in LH03. You will be given the full class time: 75 minutes. Use it wisely! Many of the roblems have short answers; try to find shortcuts. You

More information

Convex Optimization methods for Computing Channel Capacity

Convex Optimization methods for Computing Channel Capacity Convex Otimization methods for Comuting Channel Caacity Abhishek Sinha Laboratory for Information and Decision Systems (LIDS), MIT sinhaa@mit.edu May 15, 2014 We consider a classical comutational roblem

More information

Analysis of some entrance probabilities for killed birth-death processes

Analysis of some entrance probabilities for killed birth-death processes Analysis of some entrance robabilities for killed birth-death rocesses Master s Thesis O.J.G. van der Velde Suervisor: Dr. F.M. Sieksma July 5, 207 Mathematical Institute, Leiden University Contents Introduction

More information

Energy Harvesting Multiple Access Channel with Peak Temperature Constraints

Energy Harvesting Multiple Access Channel with Peak Temperature Constraints Energy Harvesting Multiple Access Channel with Peak Temperature Constraints Abdulrahman Baknina, Omur Ozel 2, and Sennur Ulukus Department of Electrical and Computer Engineering, University of Maryland,

More information

On Code Design for Simultaneous Energy and Information Transfer

On Code Design for Simultaneous Energy and Information Transfer On Code Design for Simultaneous Energy and Information Transfer Anshoo Tandon Electrical and Comuter Engineering National University of Singaore Email: anshoo@nus.edu.sg Mehul Motani Electrical and Comuter

More information

Homework Set #3 Rates definitions, Channel Coding, Source-Channel coding

Homework Set #3 Rates definitions, Channel Coding, Source-Channel coding Homework Set # Rates definitions, Channel Coding, Source-Channel coding. Rates (a) Channels coding Rate: Assuming you are sending 4 different messages using usages of a channel. What is the rate (in bits

More information

Can Feedback Increase the Capacity of the Energy Harvesting Channel?

Can Feedback Increase the Capacity of the Energy Harvesting Channel? Can Feedback Increase the Capacity of the Energy Harvesting Channel? Dor Shaviv EE Dept., Stanford University shaviv@stanford.edu Ayfer Özgür EE Dept., Stanford University aozgur@stanford.edu Haim Permuter

More information

Optimal Random Access and Random Spectrum Sensing for an Energy Harvesting Cognitive Radio with and without Primary Feedback Leveraging

Optimal Random Access and Random Spectrum Sensing for an Energy Harvesting Cognitive Radio with and without Primary Feedback Leveraging 1 Otimal Random Access and Random Sectrum Sensing for an Energy Harvesting Cognitive Radio with and without Primary Feedback Leveraging Ahmed El Shafie, Member, IEEE, arxiv:1401.0340v3 [cs.it] 27 Ar 2014

More information

Age of Information: Whittle Index for Scheduling Stochastic Arrivals

Age of Information: Whittle Index for Scheduling Stochastic Arrivals Age of Information: Whittle Index for Scheduling Stochastic Arrivals Yu-Pin Hsu Deartment of Communication Engineering National Taiei University yuinhsu@mail.ntu.edu.tw arxiv:80.03422v2 [math.oc] 7 Ar

More information

Elementary Analysis in Q p

Elementary Analysis in Q p Elementary Analysis in Q Hannah Hutter, May Szedlák, Phili Wirth November 17, 2011 This reort follows very closely the book of Svetlana Katok 1. 1 Sequences and Series In this section we will see some

More information

Coding Along Hermite Polynomials for Gaussian Noise Channels

Coding Along Hermite Polynomials for Gaussian Noise Channels Coding Along Hermite olynomials for Gaussian Noise Channels Emmanuel A. Abbe IG, EFL Lausanne, 1015 CH Email: emmanuel.abbe@efl.ch Lizhong Zheng LIDS, MIT Cambridge, MA 0139 Email: lizhong@mit.edu Abstract

More information

4. Score normalization technical details We now discuss the technical details of the score normalization method.

4. Score normalization technical details We now discuss the technical details of the score normalization method. SMT SCORING SYSTEM This document describes the scoring system for the Stanford Math Tournament We begin by giving an overview of the changes to scoring and a non-technical descrition of the scoring rules

More information

s v 0 q 0 v 1 q 1 v 2 (q 2) v 3 q 3 v 4

s v 0 q 0 v 1 q 1 v 2 (q 2) v 3 q 3 v 4 Discrete Adative Transmission for Fading Channels Lang Lin Λ, Roy D. Yates, Predrag Sasojevic WINLAB, Rutgers University 7 Brett Rd., NJ- fllin, ryates, sasojevg@winlab.rutgers.edu Abstract In this work

More information

Training sequence optimization for frequency selective channels with MAP equalization

Training sequence optimization for frequency selective channels with MAP equalization 532 ISCCSP 2008, Malta, 12-14 March 2008 raining sequence otimization for frequency selective channels with MAP equalization Imed Hadj Kacem, Noura Sellami Laboratoire LEI ENIS, Route Sokra km 35 BP 3038

More information

Sampling and Distortion Tradeoffs for Bandlimited Periodic Signals

Sampling and Distortion Tradeoffs for Bandlimited Periodic Signals Samling and Distortion radeoffs for Bandlimited Periodic Signals Elaheh ohammadi and Farokh arvasti Advanced Communications Research Institute ACRI Deartment of Electrical Engineering Sharif University

More information

On the capacity of the general trapdoor channel with feedback

On the capacity of the general trapdoor channel with feedback On the caacity of the general tradoor channel with feedback Jui Wu and Achilleas Anastasooulos Electrical Engineering and Comuter Science Deartment University of Michigan Ann Arbor, MI, 48109-1 email:

More information

Analysis of Multi-Hop Emergency Message Propagation in Vehicular Ad Hoc Networks

Analysis of Multi-Hop Emergency Message Propagation in Vehicular Ad Hoc Networks Analysis of Multi-Ho Emergency Message Proagation in Vehicular Ad Hoc Networks ABSTRACT Vehicular Ad Hoc Networks (VANETs) are attracting the attention of researchers, industry, and governments for their

More information

Approximating min-max k-clustering

Approximating min-max k-clustering Aroximating min-max k-clustering Asaf Levin July 24, 2007 Abstract We consider the roblems of set artitioning into k clusters with minimum total cost and minimum of the maximum cost of a cluster. The cost

More information

Interactive Hypothesis Testing Against Independence

Interactive Hypothesis Testing Against Independence 013 IEEE International Symosium on Information Theory Interactive Hyothesis Testing Against Indeendence Yu Xiang and Young-Han Kim Deartment of Electrical and Comuter Engineering University of California,

More information

MATHEMATICAL MODELLING OF THE WIRELESS COMMUNICATION NETWORK

MATHEMATICAL MODELLING OF THE WIRELESS COMMUNICATION NETWORK Comuter Modelling and ew Technologies, 5, Vol.9, o., 3-39 Transort and Telecommunication Institute, Lomonosov, LV-9, Riga, Latvia MATHEMATICAL MODELLIG OF THE WIRELESS COMMUICATIO ETWORK M. KOPEETSK Deartment

More information

WIRELESS energy transfer (WET), where receivers harvest

WIRELESS energy transfer (WET), where receivers harvest 6898 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 10, OCTOBER 2016 User-Centric Energy Efficiency Maximization for Wireless Powered Communications Qingqing Wu, Student Member, IEEE, Wen Chen,

More information

Optimal Random Access and Random Spectrum Sensing for an Energy Harvesting Cognitive Radio with and without Primary Feedback Leveraging

Optimal Random Access and Random Spectrum Sensing for an Energy Harvesting Cognitive Radio with and without Primary Feedback Leveraging Otimal Random Access and Random Sectrum Sensing for an Energy Harvesting Cognitive Radio with and without Primary Feedback Leveraging Ahmed El Shafie Wireless Intelligent Networks Center WINC, Nile University,

More information

Various Proofs for the Decrease Monotonicity of the Schatten s Power Norm, Various Families of R n Norms and Some Open Problems

Various Proofs for the Decrease Monotonicity of the Schatten s Power Norm, Various Families of R n Norms and Some Open Problems Int. J. Oen Problems Comt. Math., Vol. 3, No. 2, June 2010 ISSN 1998-6262; Coyright c ICSRS Publication, 2010 www.i-csrs.org Various Proofs for the Decrease Monotonicity of the Schatten s Power Norm, Various

More information

Anytime communication over the Gilbert-Eliot channel with noiseless feedback

Anytime communication over the Gilbert-Eliot channel with noiseless feedback Anytime communication over the Gilbert-Eliot channel with noiseless feedback Anant Sahai, Salman Avestimehr, Paolo Minero Deartment of Electrical Engineering and Comuter Sciences University of California

More information

Linear diophantine equations for discrete tomography

Linear diophantine equations for discrete tomography Journal of X-Ray Science and Technology 10 001 59 66 59 IOS Press Linear diohantine euations for discrete tomograhy Yangbo Ye a,gewang b and Jiehua Zhu a a Deartment of Mathematics, The University of Iowa,

More information

Energy State Amplification in an Energy Harvesting Communication System

Energy State Amplification in an Energy Harvesting Communication System Energy State Amplification in an Energy Harvesting Communication System Omur Ozel Sennur Ulukus Department of Electrical and Computer Engineering University of Maryland College Park, MD 20742 omur@umd.edu

More information

An Analysis of TCP over Random Access Satellite Links

An Analysis of TCP over Random Access Satellite Links An Analysis of over Random Access Satellite Links Chunmei Liu and Eytan Modiano Massachusetts Institute of Technology Cambridge, MA 0239 Email: mayliu, modiano@mit.edu Abstract This aer analyzes the erformance

More information

State Estimation with ARMarkov Models

State Estimation with ARMarkov Models Deartment of Mechanical and Aerosace Engineering Technical Reort No. 3046, October 1998. Princeton University, Princeton, NJ. State Estimation with ARMarkov Models Ryoung K. Lim 1 Columbia University,

More information

Online Appendix to Accompany AComparisonof Traditional and Open-Access Appointment Scheduling Policies

Online Appendix to Accompany AComparisonof Traditional and Open-Access Appointment Scheduling Policies Online Aendix to Accomany AComarisonof Traditional and Oen-Access Aointment Scheduling Policies Lawrence W. Robinson Johnson Graduate School of Management Cornell University Ithaca, NY 14853-6201 lwr2@cornell.edu

More information

Radial Basis Function Networks: Algorithms

Radial Basis Function Networks: Algorithms Radial Basis Function Networks: Algorithms Introduction to Neural Networks : Lecture 13 John A. Bullinaria, 2004 1. The RBF Maing 2. The RBF Network Architecture 3. Comutational Power of RBF Networks 4.

More information

Research Article An iterative Algorithm for Hemicontractive Mappings in Banach Spaces

Research Article An iterative Algorithm for Hemicontractive Mappings in Banach Spaces Abstract and Alied Analysis Volume 2012, Article ID 264103, 11 ages doi:10.1155/2012/264103 Research Article An iterative Algorithm for Hemicontractive Maings in Banach Saces Youli Yu, 1 Zhitao Wu, 2 and

More information

On the Role of Finite Queues in Cooperative Cognitive Radio Networks with Energy Harvesting

On the Role of Finite Queues in Cooperative Cognitive Radio Networks with Energy Harvesting On the Role of Finite Queues in Cooerative Cognitive Radio Networks with Energy Harvesting Mohamed A. Abd-Elmagid, Tamer Elatt, and Karim G. Seddik Wireless Intelligent Networks Center (WINC), Nile University,

More information

CHAPTER-II Control Charts for Fraction Nonconforming using m-of-m Runs Rules

CHAPTER-II Control Charts for Fraction Nonconforming using m-of-m Runs Rules CHAPTER-II Control Charts for Fraction Nonconforming using m-of-m Runs Rules. Introduction: The is widely used in industry to monitor the number of fraction nonconforming units. A nonconforming unit is

More information

Some Unitary Space Time Codes From Sphere Packing Theory With Optimal Diversity Product of Code Size

Some Unitary Space Time Codes From Sphere Packing Theory With Optimal Diversity Product of Code Size IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 5, NO., DECEMBER 4 336 Some Unitary Sace Time Codes From Shere Packing Theory With Otimal Diversity Product of Code Size Haiquan Wang, Genyuan Wang, and Xiang-Gen

More information

Event-triggering stabilization of complex linear systems with disturbances over digital channels

Event-triggering stabilization of complex linear systems with disturbances over digital channels Event-triggering stabilization of comlex linear systems with disturbances over digital channels Mohammad Javad Khojasteh, Mojtaba Hedayatour, Jorge Cortés, Massimo Franceschetti Abstract As stos and auses

More information

Achieving Shannon Capacity Region as Secrecy Rate Region in a Multiple Access Wiretap Channel

Achieving Shannon Capacity Region as Secrecy Rate Region in a Multiple Access Wiretap Channel Achieving Shannon Capacity Region as Secrecy Rate Region in a Multiple Access Wiretap Channel Shahid Mehraj Shah and Vinod Sharma Department of Electrical Communication Engineering, Indian Institute of

More information

The non-stochastic multi-armed bandit problem

The non-stochastic multi-armed bandit problem Submitted for journal ublication. The non-stochastic multi-armed bandit roblem Peter Auer Institute for Theoretical Comuter Science Graz University of Technology A-8010 Graz (Austria) auer@igi.tu-graz.ac.at

More information

Amin, Osama; Abediseid, Walid; Alouini, Mohamed-Slim. Institute of Electrical and Electronics Engineers (IEEE)

Amin, Osama; Abediseid, Walid; Alouini, Mohamed-Slim. Institute of Electrical and Electronics Engineers (IEEE) KAUST Reository Outage erformance of cognitive radio systems with Imroer Gaussian signaling Item tye Authors Erint version DOI Publisher Journal Rights Conference Paer Amin Osama; Abediseid Walid; Alouini

More information

Asymptotically Optimal Simulation Allocation under Dependent Sampling

Asymptotically Optimal Simulation Allocation under Dependent Sampling Asymtotically Otimal Simulation Allocation under Deendent Samling Xiaoing Xiong The Robert H. Smith School of Business, University of Maryland, College Park, MD 20742-1815, USA, xiaoingx@yahoo.com Sandee

More information

Robust Predictive Control of Input Constraints and Interference Suppression for Semi-Trailer System

Robust Predictive Control of Input Constraints and Interference Suppression for Semi-Trailer System Vol.7, No.7 (4),.37-38 htt://dx.doi.org/.457/ica.4.7.7.3 Robust Predictive Control of Inut Constraints and Interference Suression for Semi-Trailer System Zhao, Yang Electronic and Information Technology

More information

t 0 Xt sup X t p c p inf t 0

t 0 Xt sup X t p c p inf t 0 SHARP MAXIMAL L -ESTIMATES FOR MARTINGALES RODRIGO BAÑUELOS AND ADAM OSȨKOWSKI ABSTRACT. Let X be a suermartingale starting from 0 which has only nonnegative jums. For each 0 < < we determine the best

More information

Broadcasting with a Battery Limited Energy Harvesting Rechargeable Transmitter

Broadcasting with a Battery Limited Energy Harvesting Rechargeable Transmitter roadcasting with a attery Limited Energy Harvesting Rechargeable Transmitter Omur Ozel, Jing Yang 2, and Sennur Ulukus Department of Electrical and Computer Engineering, University of Maryland, College

More information

Capacity of channel with energy harvesting transmitter

Capacity of channel with energy harvesting transmitter IET Communications Research Article Capacity of channel with energy harvesting transmitter ISSN 75-868 Received on nd May 04 Accepted on 7th October 04 doi: 0.049/iet-com.04.0445 www.ietdl.org Hamid Ghanizade

More information

MODEL-BASED MULTIPLE FAULT DETECTION AND ISOLATION FOR NONLINEAR SYSTEMS

MODEL-BASED MULTIPLE FAULT DETECTION AND ISOLATION FOR NONLINEAR SYSTEMS MODEL-BASED MULIPLE FAUL DEECION AND ISOLAION FOR NONLINEAR SYSEMS Ivan Castillo, and homas F. Edgar he University of exas at Austin Austin, X 78712 David Hill Chemstations Houston, X 77009 Abstract A

More information

Shadow Computing: An Energy-Aware Fault Tolerant Computing Model

Shadow Computing: An Energy-Aware Fault Tolerant Computing Model Shadow Comuting: An Energy-Aware Fault Tolerant Comuting Model Bryan Mills, Taieb Znati, Rami Melhem Deartment of Comuter Science University of Pittsburgh (bmills, znati, melhem)@cs.itt.edu Index Terms

More information

Distributed Rule-Based Inference in the Presence of Redundant Information

Distributed Rule-Based Inference in the Presence of Redundant Information istribution Statement : roved for ublic release; distribution is unlimited. istributed Rule-ased Inference in the Presence of Redundant Information June 8, 004 William J. Farrell III Lockheed Martin dvanced

More information

A continuous review inventory model with the controllable production rate of the manufacturer

A continuous review inventory model with the controllable production rate of the manufacturer Intl. Trans. in O. Res. 12 (2005) 247 258 INTERNATIONAL TRANSACTIONS IN OERATIONAL RESEARCH A continuous review inventory model with the controllable roduction rate of the manufacturer I. K. Moon and B.

More information

Distributed K-means over Compressed Binary Data

Distributed K-means over Compressed Binary Data 1 Distributed K-means over Comressed Binary Data Elsa DUPRAZ Telecom Bretagne; UMR CNRS 6285 Lab-STICC, Brest, France arxiv:1701.03403v1 [cs.it] 12 Jan 2017 Abstract We consider a networ of binary-valued

More information

1-way quantum finite automata: strengths, weaknesses and generalizations

1-way quantum finite automata: strengths, weaknesses and generalizations 1-way quantum finite automata: strengths, weaknesses and generalizations arxiv:quant-h/9802062v3 30 Se 1998 Andris Ambainis UC Berkeley Abstract Rūsiņš Freivalds University of Latvia We study 1-way quantum

More information

arxiv: v1 [physics.data-an] 26 Oct 2012

arxiv: v1 [physics.data-an] 26 Oct 2012 Constraints on Yield Parameters in Extended Maximum Likelihood Fits Till Moritz Karbach a, Maximilian Schlu b a TU Dortmund, Germany, moritz.karbach@cern.ch b TU Dortmund, Germany, maximilian.schlu@cern.ch

More information

Estimation of the large covariance matrix with two-step monotone missing data

Estimation of the large covariance matrix with two-step monotone missing data Estimation of the large covariance matrix with two-ste monotone missing data Masashi Hyodo, Nobumichi Shutoh 2, Takashi Seo, and Tatjana Pavlenko 3 Deartment of Mathematical Information Science, Tokyo

More information

DEPARTMENT OF ECONOMICS ISSN DISCUSSION PAPER 20/07 TWO NEW EXPONENTIAL FAMILIES OF LORENZ CURVES

DEPARTMENT OF ECONOMICS ISSN DISCUSSION PAPER 20/07 TWO NEW EXPONENTIAL FAMILIES OF LORENZ CURVES DEPARTMENT OF ECONOMICS ISSN 1441-549 DISCUSSION PAPER /7 TWO NEW EXPONENTIAL FAMILIES OF LORENZ CURVES ZuXiang Wang * & Russell Smyth ABSTRACT We resent two new Lorenz curve families by using the basic

More information

Research Article A New Method to Study Analytic Inequalities

Research Article A New Method to Study Analytic Inequalities Hindawi Publishing Cororation Journal of Inequalities and Alications Volume 200, Article ID 69802, 3 ages doi:0.55/200/69802 Research Article A New Method to Study Analytic Inequalities Xiao-Ming Zhang

More information

Feedback-error control

Feedback-error control Chater 4 Feedback-error control 4.1 Introduction This chater exlains the feedback-error (FBE) control scheme originally described by Kawato [, 87, 8]. FBE is a widely used neural network based controller

More information

GRACEFUL NUMBERS. KIRAN R. BHUTANI and ALEXANDER B. LEVIN. Received 14 May 2001

GRACEFUL NUMBERS. KIRAN R. BHUTANI and ALEXANDER B. LEVIN. Received 14 May 2001 IJMMS 29:8 2002 495 499 PII S06720200765 htt://immshindawicom Hindawi Publishing Cor GRACEFUL NUMBERS KIRAN R BHUTANI and ALEXANDER B LEVIN Received 4 May 200 We construct a labeled grah Dn that reflects

More information

System Reliability Estimation and Confidence Regions from Subsystem and Full System Tests

System Reliability Estimation and Confidence Regions from Subsystem and Full System Tests 009 American Control Conference Hyatt Regency Riverfront, St. Louis, MO, USA June 0-, 009 FrB4. System Reliability Estimation and Confidence Regions from Subsystem and Full System Tests James C. Sall Abstract

More information

The decision-feedback equalizer optimization for Gaussian noise

The decision-feedback equalizer optimization for Gaussian noise Journal of Theoretical and Alied Comuter Science Vol. 8 No. 4 4. 5- ISSN 99-634 (rinted 3-5653 (online htt://www.jtacs.org The decision-feedback eualizer otimization for Gaussian noise Arkadiusz Grzbowski

More information

MODELING THE RELIABILITY OF C4ISR SYSTEMS HARDWARE/SOFTWARE COMPONENTS USING AN IMPROVED MARKOV MODEL

MODELING THE RELIABILITY OF C4ISR SYSTEMS HARDWARE/SOFTWARE COMPONENTS USING AN IMPROVED MARKOV MODEL Technical Sciences and Alied Mathematics MODELING THE RELIABILITY OF CISR SYSTEMS HARDWARE/SOFTWARE COMPONENTS USING AN IMPROVED MARKOV MODEL Cezar VASILESCU Regional Deartment of Defense Resources Management

More information

#A64 INTEGERS 18 (2018) APPLYING MODULAR ARITHMETIC TO DIOPHANTINE EQUATIONS

#A64 INTEGERS 18 (2018) APPLYING MODULAR ARITHMETIC TO DIOPHANTINE EQUATIONS #A64 INTEGERS 18 (2018) APPLYING MODULAR ARITHMETIC TO DIOPHANTINE EQUATIONS Ramy F. Taki ElDin Physics and Engineering Mathematics Deartment, Faculty of Engineering, Ain Shams University, Cairo, Egyt

More information

Periodic scheduling 05/06/

Periodic scheduling 05/06/ Periodic scheduling T T or eriodic scheduling, the best that we can do is to design an algorithm which will always find a schedule if one exists. A scheduler is defined to be otimal iff it will find a

More information

Formal Modeling in Cognitive Science Lecture 29: Noisy Channel Model and Applications;

Formal Modeling in Cognitive Science Lecture 29: Noisy Channel Model and Applications; Formal Modeling in Cognitive Science Lecture 9: and ; ; Frank Keller School of Informatics University of Edinburgh keller@inf.ed.ac.uk Proerties of 3 March, 6 Frank Keller Formal Modeling in Cognitive

More information

Sum Capacity of General Deterministic Interference Channel with Channel Output Feedback

Sum Capacity of General Deterministic Interference Channel with Channel Output Feedback Sum Capacity of General Deterministic Interference Channel with Channel Output Feedback Achaleshwar Sahai Department of ECE, Rice University, Houston, TX 775. as27@rice.edu Vaneet Aggarwal Department of

More information

Notes on Instrumental Variables Methods

Notes on Instrumental Variables Methods Notes on Instrumental Variables Methods Michele Pellizzari IGIER-Bocconi, IZA and frdb 1 The Instrumental Variable Estimator Instrumental variable estimation is the classical solution to the roblem of

More information

COMPARISON OF VARIOUS OPTIMIZATION TECHNIQUES FOR DESIGN FIR DIGITAL FILTERS

COMPARISON OF VARIOUS OPTIMIZATION TECHNIQUES FOR DESIGN FIR DIGITAL FILTERS NCCI 1 -National Conference on Comutational Instrumentation CSIO Chandigarh, INDIA, 19- March 1 COMPARISON OF VARIOUS OPIMIZAION ECHNIQUES FOR DESIGN FIR DIGIAL FILERS Amanjeet Panghal 1, Nitin Mittal,Devender

More information

A MIXED CONTROL CHART ADAPTED TO THE TRUNCATED LIFE TEST BASED ON THE WEIBULL DISTRIBUTION

A MIXED CONTROL CHART ADAPTED TO THE TRUNCATED LIFE TEST BASED ON THE WEIBULL DISTRIBUTION O P E R A T I O N S R E S E A R C H A N D D E C I S I O N S No. 27 DOI:.5277/ord73 Nasrullah KHAN Muhammad ASLAM 2 Kyung-Jun KIM 3 Chi-Hyuck JUN 4 A MIXED CONTROL CHART ADAPTED TO THE TRUNCATED LIFE TEST

More information

For q 0; 1; : : : ; `? 1, we have m 0; 1; : : : ; q? 1. The set fh j(x) : j 0; 1; ; : : : ; `? 1g forms a basis for the tness functions dened on the i

For q 0; 1; : : : ; `? 1, we have m 0; 1; : : : ; q? 1. The set fh j(x) : j 0; 1; ; : : : ; `? 1g forms a basis for the tness functions dened on the i Comuting with Haar Functions Sami Khuri Deartment of Mathematics and Comuter Science San Jose State University One Washington Square San Jose, CA 9519-0103, USA khuri@juiter.sjsu.edu Fax: (40)94-500 Keywords:

More information

Lower Confidence Bound for Process-Yield Index S pk with Autocorrelated Process Data

Lower Confidence Bound for Process-Yield Index S pk with Autocorrelated Process Data Quality Technology & Quantitative Management Vol. 1, No.,. 51-65, 15 QTQM IAQM 15 Lower onfidence Bound for Process-Yield Index with Autocorrelated Process Data Fu-Kwun Wang * and Yeneneh Tamirat Deartment

More information

Research Article Controllability of Linear Discrete-Time Systems with Both Delayed States and Delayed Inputs

Research Article Controllability of Linear Discrete-Time Systems with Both Delayed States and Delayed Inputs Abstract and Alied Analysis Volume 203 Article ID 97546 5 ages htt://dxdoiorg/055/203/97546 Research Article Controllability of Linear Discrete-Time Systems with Both Delayed States and Delayed Inuts Hong

More information

Understanding and Using Availability

Understanding and Using Availability Understanding and Using Availability Jorge Luis Romeu, Ph.D. ASQ CQE/CRE, & Senior Member Email: romeu@cortland.edu htt://myrofile.cos.com/romeu ASQ/RD Webinar Series Noviembre 5, J. L. Romeu - Consultant

More information

Combining Logistic Regression with Kriging for Mapping the Risk of Occurrence of Unexploded Ordnance (UXO)

Combining Logistic Regression with Kriging for Mapping the Risk of Occurrence of Unexploded Ordnance (UXO) Combining Logistic Regression with Kriging for Maing the Risk of Occurrence of Unexloded Ordnance (UXO) H. Saito (), P. Goovaerts (), S. A. McKenna (2) Environmental and Water Resources Engineering, Deartment

More information

A Coordinate System for Gaussian Networks

A Coordinate System for Gaussian Networks A Coordinate System for Gaussian Networs The MIT Faculty has made this article oenly available. Please share how this access benefits you. Your story matters. Citation As Published Publisher Abbe, Emmanuel,

More information

An Investigation on the Numerical Ill-conditioning of Hybrid State Estimators

An Investigation on the Numerical Ill-conditioning of Hybrid State Estimators An Investigation on the Numerical Ill-conditioning of Hybrid State Estimators S. K. Mallik, Student Member, IEEE, S. Chakrabarti, Senior Member, IEEE, S. N. Singh, Senior Member, IEEE Deartment of Electrical

More information

Topic: Lower Bounds on Randomized Algorithms Date: September 22, 2004 Scribe: Srinath Sridhar

Topic: Lower Bounds on Randomized Algorithms Date: September 22, 2004 Scribe: Srinath Sridhar 15-859(M): Randomized Algorithms Lecturer: Anuam Guta Toic: Lower Bounds on Randomized Algorithms Date: Setember 22, 2004 Scribe: Srinath Sridhar 4.1 Introduction In this lecture, we will first consider

More information

MATH 2710: NOTES FOR ANALYSIS

MATH 2710: NOTES FOR ANALYSIS MATH 270: NOTES FOR ANALYSIS The main ideas we will learn from analysis center around the idea of a limit. Limits occurs in several settings. We will start with finite limits of sequences, then cover infinite

More information

On Wald-Type Optimal Stopping for Brownian Motion

On Wald-Type Optimal Stopping for Brownian Motion J Al Probab Vol 34, No 1, 1997, (66-73) Prerint Ser No 1, 1994, Math Inst Aarhus On Wald-Tye Otimal Stoing for Brownian Motion S RAVRSN and PSKIR The solution is resented to all otimal stoing roblems of

More information

Vision Graph Construction in Wireless Multimedia Sensor Networks

Vision Graph Construction in Wireless Multimedia Sensor Networks University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln CSE Conference and Worksho Paers Comuter Science and Engineering, Deartment of 21 Vision Grah Construction in Wireless Multimedia

More information

Analyses of Orthogonal and Non-Orthogonal Steering Vectors at Millimeter Wave Systems

Analyses of Orthogonal and Non-Orthogonal Steering Vectors at Millimeter Wave Systems Analyses of Orthogonal and Non-Orthogonal Steering Vectors at Millimeter Wave Systems Hsiao-Lan Chiang, Tobias Kadur, and Gerhard Fettweis Vodafone Chair for Mobile Communications Technische Universität

More information

1. INTRODUCTION. Fn 2 = F j F j+1 (1.1)

1. INTRODUCTION. Fn 2 = F j F j+1 (1.1) CERTAIN CLASSES OF FINITE SUMS THAT INVOLVE GENERALIZED FIBONACCI AND LUCAS NUMBERS The beautiful identity R.S. Melham Deartment of Mathematical Sciences, University of Technology, Sydney PO Box 23, Broadway,

More information

John Weatherwax. Analysis of Parallel Depth First Search Algorithms

John Weatherwax. Analysis of Parallel Depth First Search Algorithms Sulementary Discussions and Solutions to Selected Problems in: Introduction to Parallel Comuting by Viin Kumar, Ananth Grama, Anshul Guta, & George Karyis John Weatherwax Chater 8 Analysis of Parallel

More information

Spatial Outage Capacity of Poisson Bipolar Networks

Spatial Outage Capacity of Poisson Bipolar Networks Satial Outage Caacity of Poisson Biolar Networks Sanket S. Kalamkar and Martin Haenggi Deartment of Electrical Engineering, University of Notre Dame, Notre Dame, IN 46556, USA E-mail: skalamka@nd.edu,

More information

Improving on the Cutset Bound via a Geometric Analysis of Typical Sets

Improving on the Cutset Bound via a Geometric Analysis of Typical Sets Imroving on the Cutset Bound via a Geometric Analysis of Tyical Sets Ayfer Özgür Stanford University CUHK, May 28, 2016 Joint work with Xiugang Wu (Stanford). Ayfer Özgür (Stanford) March 16 1 / 33 Gaussian

More information

Characterizing the Behavior of a Probabilistic CMOS Switch Through Analytical Models and Its Verification Through Simulations

Characterizing the Behavior of a Probabilistic CMOS Switch Through Analytical Models and Its Verification Through Simulations Characterizing the Behavior of a Probabilistic CMOS Switch Through Analytical Models and Its Verification Through Simulations PINAR KORKMAZ, BILGE E. S. AKGUL and KRISHNA V. PALEM Georgia Institute of

More information

Convolutional Codes. Lecture 13. Figure 93: Encoder for rate 1/2 constraint length 3 convolutional code.

Convolutional Codes. Lecture 13. Figure 93: Encoder for rate 1/2 constraint length 3 convolutional code. Convolutional Codes Goals Lecture Be able to encode using a convolutional code Be able to decode a convolutional code received over a binary symmetric channel or an additive white Gaussian channel Convolutional

More information

I - Information theory basics

I - Information theory basics I - Information theor basics Introduction To communicate, that is, to carr information between two oints, we can emlo analog or digital transmission techniques. In digital communications the message is

More information

Towards understanding the Lorenz curve using the Uniform distribution. Chris J. Stephens. Newcastle City Council, Newcastle upon Tyne, UK

Towards understanding the Lorenz curve using the Uniform distribution. Chris J. Stephens. Newcastle City Council, Newcastle upon Tyne, UK Towards understanding the Lorenz curve using the Uniform distribution Chris J. Stehens Newcastle City Council, Newcastle uon Tyne, UK (For the Gini-Lorenz Conference, University of Siena, Italy, May 2005)

More information

Understanding and Using Availability

Understanding and Using Availability Understanding and Using Availability Jorge Luis Romeu, Ph.D. ASQ CQE/CRE, & Senior Member C. Stat Fellow, Royal Statistical Society Past Director, Region II (NY & PA) Director: Juarez Lincoln Marti Int

More information

Vehicular Ad-hoc Networks using slotted Aloha: Point-to-Point, Emergency and Broadcast Communications

Vehicular Ad-hoc Networks using slotted Aloha: Point-to-Point, Emergency and Broadcast Communications Vehicular Ad-hoc Networks using slotted Aloha: Point-to-Point, Emergency and Broadcast Communications Bartłomiej Błaszczyszyn Paul Muhlethaler Nadjib Achir INRIA/ENS INRIA Rocquencourt INRIA Rocquencourt

More information

BEST POSSIBLE DENSITIES OF DICKSON m-tuples, AS A CONSEQUENCE OF ZHANG-MAYNARD-TAO

BEST POSSIBLE DENSITIES OF DICKSON m-tuples, AS A CONSEQUENCE OF ZHANG-MAYNARD-TAO BEST POSSIBLE DENSITIES OF DICKSON m-tuples, AS A CONSEQUENCE OF ZHANG-MAYNARD-TAO ANDREW GRANVILLE, DANIEL M. KANE, DIMITRIS KOUKOULOPOULOS, AND ROBERT J. LEMKE OLIVER Abstract. We determine for what

More information

Gradient Based Iterative Algorithms for Solving a Class of Matrix Equations

Gradient Based Iterative Algorithms for Solving a Class of Matrix Equations 1216 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 50, NO. 8, AUGUST 2005 Exanding (A3)gives M 0 8 T +1 T +18 =( T +1 +1 ) F = I (A4) g = 0F 8 T +1 =( T +1 +1 ) f = T +1 +1 + T +18 F 8 T +1 =( T +1 +1 )

More information

Construction of High-Girth QC-LDPC Codes

Construction of High-Girth QC-LDPC Codes MITSUBISHI ELECTRIC RESEARCH LABORATORIES htt://wwwmerlcom Construction of High-Girth QC-LDPC Codes Yige Wang, Jonathan Yedidia, Stark Draer TR2008-061 Setember 2008 Abstract We describe a hill-climbing

More information

A Public-Key Cryptosystem Based on Lucas Sequences

A Public-Key Cryptosystem Based on Lucas Sequences Palestine Journal of Mathematics Vol. 1(2) (2012), 148 152 Palestine Polytechnic University-PPU 2012 A Public-Key Crytosystem Based on Lucas Sequences Lhoussain El Fadil Communicated by Ayman Badawi MSC2010

More information

arxiv: v1 [cs.gt] 2 Nov 2018

arxiv: v1 [cs.gt] 2 Nov 2018 Tight Aroximation Ratio of Anonymous Pricing Yaonan Jin Pinyan Lu Qi Qi Zhihao Gavin Tang Tao Xiao arxiv:8.763v [cs.gt] 2 Nov 28 Abstract We consider two canonical Bayesian mechanism design settings. In

More information

Optimal Design of Truss Structures Using a Neutrosophic Number Optimization Model under an Indeterminate Environment

Optimal Design of Truss Structures Using a Neutrosophic Number Optimization Model under an Indeterminate Environment Neutrosohic Sets and Systems Vol 14 016 93 University of New Mexico Otimal Design of Truss Structures Using a Neutrosohic Number Otimization Model under an Indeterminate Environment Wenzhong Jiang & Jun

More information

On a Markov Game with Incomplete Information

On a Markov Game with Incomplete Information On a Markov Game with Incomlete Information Johannes Hörner, Dinah Rosenberg y, Eilon Solan z and Nicolas Vieille x{ January 24, 26 Abstract We consider an examle of a Markov game with lack of information

More information

Uncertainty Modeling with Interval Type-2 Fuzzy Logic Systems in Mobile Robotics

Uncertainty Modeling with Interval Type-2 Fuzzy Logic Systems in Mobile Robotics Uncertainty Modeling with Interval Tye-2 Fuzzy Logic Systems in Mobile Robotics Ondrej Linda, Student Member, IEEE, Milos Manic, Senior Member, IEEE bstract Interval Tye-2 Fuzzy Logic Systems (IT2 FLSs)

More information

q-ary Symmetric Channel for Large q

q-ary Symmetric Channel for Large q List-Message Passing Achieves Caacity on the q-ary Symmetric Channel for Large q Fan Zhang and Henry D Pfister Deartment of Electrical and Comuter Engineering, Texas A&M University {fanzhang,hfister}@tamuedu

More information

AN IMPROVED BABY-STEP-GIANT-STEP METHOD FOR CERTAIN ELLIPTIC CURVES. 1. Introduction

AN IMPROVED BABY-STEP-GIANT-STEP METHOD FOR CERTAIN ELLIPTIC CURVES. 1. Introduction J. Al. Math. & Comuting Vol. 20(2006), No. 1-2,. 485-489 AN IMPROVED BABY-STEP-GIANT-STEP METHOD FOR CERTAIN ELLIPTIC CURVES BYEONG-KWEON OH, KIL-CHAN HA AND JANGHEON OH Abstract. In this aer, we slightly

More information