Optimal Resource Allocation for Multi-User MEC with Arbitrary Task Arrival Times and Deadlines

Size: px
Start display at page:

Download "Optimal Resource Allocation for Multi-User MEC with Arbitrary Task Arrival Times and Deadlines"

Transcription

1 Optimal Resource Allocation for Multi-User MEC with Arbitrary Task Arrival Times and Deadlines Xinyun Wang, Ying Cui, Zhi Liu, Junfeng Guo, Mingyu Yang Abstract In this paper, we would like to investigate optimal resource allocation for a more practical multiuser mobile edge computing (MEC) system. First, we consider a computation task model with nonnegligible sizes of computation results and arbitrary task arrival times and deadlines. Based on it, we further establish a computation offloading model considering non-negligible execution durations and allowing parallel transmissions and executions for different tasks. Then, we formulate the weighted sum energy consumption minimization problem to optimize the task operation sequences and starting times for uploading, executing and downloading as well as uploading and downloading time durations. The problem is a challenging mixed discrete-continuous optimization problem. By analyzing its structural properties, we develop an algorithm to obtain an optimal solution. In addition, using several optimization techniques, we transform the problem to an equivalent Difference of Convex (DC) problem, and develop a low-complexity algorithm to obtain a suboptimal solution using Penalty Convex Concave Procedure (CCP). Both analytical and numerical results demonstrate the importance of the optimization of operation sequences in multi-user MEC systems, which is usually neglected in existing studies based on simplified models for multi-user MEC systems. I. INTRODUCTION With the support of megapixel on-device cameras and hi-precision built-in sensors, many advanced mobile applications, e.g., augmented reality, interactive online gaming and multimedia transformation are emerging. These desktop-level applications are typically computationintensive and latency-sensitive, posing increasingly high burdens on resource-limited mobile X. Wang, Y. Cui, J. Guo and M. Yang are with Shanghai Jiao Tong University, China. Z. Liu is with Shizuoka University, Japan.

2 2 devices. Mobile edge computing (MEC) is a promising technology that provides powerful computing capability at the wireless edge to improve the quality of experience of mobile users for these advanced applications. In an MEC system, computation tasks of mobile users can be offloaded to a serving node (e.g., base station and access point). Designing efficient MEC systems requires a joint optimization of communication and computation resources among distributed mobiles and MEC servers. A significant amount of research effort has been devoted to energy-efficient resource allocation for MEC systems [1] [8]. For example, [2] [4] consider energy-efficient resource allocation, such as offloading control and transmission power and time duration allocation, for computation tasks with the same arrival time and deadline. However, the synchronous assumption for computation tasks is suitable only for a limited number of applications. To obtain more practical energy-efficient resource allocation, [5], [6], [8] consider different task arrival times and deadlines. Note that [4], [5], [8] assume that the sizes of computation results are negligible, and fail to take account of the resource consumption for transmitting computation results from a serving node to mobiles. This assumption may not hold for many applications with computation results of large sizes, such as augmented reality, interactive online gaming and multi-media transformation. In addition, [2], [6], [8] assume that task execution durations are negligible. This assumption may not be suitable for several applications with heavy computation loads and long execution durations, such as multi-media transformation and 3D modeling/rendering. Although [1], [3] consider non-negligible task execution durations and non-negligible computation result sizes, they ignore the fact that the execution of one task can be conducted during the transmission of another task (i.e, transmissions and executions for different tasks can be performed in parallel). Under the simplifying assumptions, [1] [6] do not consider the optimization of operation sequences, which significantly reduces the opportunities for parallel processing. Notice that these opportunities may lead to great delay reduction under power constraints or energy consumption reduction under deadline constraints, especially for tasks with large sizes of computation results and long execution durations. In our previous work [7], we consider non-negligible computation result sizes and non-negligible task execution durations, and allow transmissions and executions of different tasks to be conducted in parallel. However, the synchronous assumption for task arrivals and completions in [7] limits the applications of the proposed solution in practical MEC systems. Therefore, further studies are required to obtain efficient resource allocation for more practical MEC systems.

3 3 In this paper, our primary goal is to extend the task model and computation offloading model for the synchronous task scenario in our previous work [7] to the general scenario with arbitrary task arrival times and deadlines, to make them more applicable. Note that this is a highly nontrivial extension, as the operation mechanisms of MEC systems will change dramatically in the general scenario. We consider a multi-user MEC system with one serving node. In particular, we consider a computation task model with non-negligible sizes of computation results and arbitrary task arrival times and deadlines. Based on it, we further establish a computation offloading model considering non-negligible execution durations and allowing parallel transmissions and executions for different tasks. Then, we formulate the weighted sum energy consumption minimization problem to optimize the task operation sequences and starting times for uploading, executing and downloading as well as uploading and downloading time durations. The problem is a challenging mixed discrete-continuous optimization problem and is NP-hard in general. By analyzing its structural properties, we obtain an equivalent formulation which separates the discrete part and the continuous part, and develop an algorithm to obtain an optimal solution. In addition, using several optimization techniques, we transform the problem to an equivalent Difference of Convex (DC) problem, and develop a low-complexity algorithm to obtain a suboptimal solution using Penalty Convex Concave Procedure (CCP) [9]. Finally, numerical results demonstrate the advantages of the proposed suboptimal solution over some optimized schemes with simple designs of operation sequences. Both analytical and numerical results in this paper demonstrate the importance of the optimization of operation sequences in designing efficient multi-user MEC systems. II. SYSTEM MODEL As illustrated in Fig. 1, we consider a multi-user MEC system consisting of one single-antenna serving node and K single-antenna users, denoted by set K {1, 2,..., K}. The serving node has powerful computing capability by running an MEC server of a constant CPU-cycle frequency (in number of CPU-cycles per second) at the network edge. Each user k K has one computationintensive and latency-sensitive (computation) task, which is offloaded to the serving node for executing. 1 We consider a computation task model with non-negligible sizes of computation results and arbitrary task arrival times and deadlines. We also establish a computation offloading model considering non-negligible task execution durations and allowing parallel transmissions 1 The optimization results obtained in this paper can be extended to study a more general scenario, where some tasks can be executed locally.

4 4 Fig. 1. System model. and executions for different tasks. The two models are more complicated but more practical than the existing ones with the same task arrival time and deadline [2] [4], [7], negligible computation result sizes [4], [5], [8], or negligible task execution durations [2], [6], [8]. A. Computation Task Model We extend the computation task model in [7] by allowing computation tasks to have different arrival times and deadlines. The computation task at user k K, referred to as task k, is characterized by five parameters, i.e., the size of the uploaded task before computation L u,k > 0 (in bits), workload L e,k > 0 (in number of CPU-cycles), size of the downloaded computation result L d,k > 0 (in bits), arrival time A k 0 and deadline D k > A k. 2 Note that the considered task model is applicable for applications with computation results of large sizes, such as augmented reality, interactive online gaming and multi-media transformation. In addition, note that for all k K, L u,k, L e,k and L d,k are determined by the nature of task k itself, and for some computation tasks, such as html2text and x264 video encoding, these three parameters can be estimated to certain extent based on some prior offline measurements. B. Computation Offloading Model We offload each computation-intensive and latency-sensitive task to the serving node for executing. Offloading task k to the serving node for executing comprises three sequential stages: uploading task k from user k to the serving node, executing task k at the serving node, and 2 In this paper, we assume A k, k K and D k, k K are known in advance, and focus on the offline scenario to obtain first-order design insights as in [8]. In our future work, we shall study the online scenario where A k, k K and D k, k K are not known, based on the results obtained for the offline scenario in this paper.

5 5 downloading the computation result from the serving node to user k. We consider Time Division Multiple Access (TDMA) in Time-Division Duplexing (TDD) mode for transmission [7], [8]. Let t u,k, t e,k and t d,k denote the uploading, execution and downloading durations (in seconds) in the three stages, respectively, where t u,k, t d,k 0, k K, (1) t e,k = L e,k, k K. (2) F Here, F denotes the fixed CPU-cycle frequency of the MEC server at the serving node. Note that t e,k is fixed, whereas t u,k and t d,k can be optimized. Denote t u (t u,k ) k K, t e (t e,k ) k K, t d (t d,k ) k K and t (t u, t d ). The considered computation offloading model is suitable for applications with heavy computation loads, such as multi-media transformation and 3D modeling/rendering. Then, we introduce new notations and constraints to mathematically specify this model. Let s u,k, s e,k and s d,k denote the starting times for uploading, executing and downloading task k, respectively. Denote s u (s u,k ) k K, s e (s e,k ) k K, s d (s d,k ) k K, s (s u, s e, s d ), and O {u, e, d}. As each of the three stages cannot be interrupted, the completion times for uploading, executing and downloading task k are given by s o,k + t o,k, o O, k K. To ensure that the uploading, execution and downloading operations of each task are conducted sequentially, we require: s e,k s u,k + t u,k, k K, (3a) s d,k s e,k + t e,k, k K. (3b) To guarantee that the uploading of each task starts after its arrival and that the downloading of each task is completed before its deadline, we have: s u,k A k, k K, (4) s d,k + t d,k D k, k K. (5) Let Q denote the set of the K! different permutations of 1, 2,..., K. Let vectors q u (q u,i ) i K, q e (q e,i ) i K and q d (q d,i ) i K denote the three operation sequences (orders) for uploading

6 6 and executing the K tasks, and downloading their computation results, respectively, where q o Q, o O, (6) and q o,i represents the i-th element of q o for all o O and i K. Denote q (q u, q e, q d ) and K K \ {K}. Given the operation sequence q o, we have the following constraints: s o,qo,i+1 s o,qo,i + t o,qo,i, o O, i K. (7) As we consider TDMA in TDD mode, at any time, there is at most one task being uploaded or downloaded. Let x i denote the number of downloading operations between the i-th uploading operation and the (i + 1)-th uploading operation, for all i K, and let x K denote the number of downloading operations after the K-th uploading operation, where x i {0} K, i K. (8) Denote x (x i ) i K. As the total number of downloading operations is K, we have: K x i = K. (9) Note that q and x jointly determine the order of all operations. Note that unlike [7], we allow q o, o O to be different and consider x, so as not to lose optimality in the case of different task arrival times and deadlines, as illustrated in Fig. 2. To ensure that at any time, there is at most one task being transmitted, it is sufficient to require that any adjacent uploading and downloading operations do not overlap in time given the constraints in (7), (8) and (9), i.e., ) I [x i 0] (s d,qd,ij=1 + t xj d,qd,ij=1 s u,qu,i+1, i K, (10a) xj I [x i+1 0] ( s u,qu,i+1 + t u,qu,i+1 ) sd,qd, ij=1 x j +1, i K, s u,qu,1 + t u,qu,1 s d,qd,1, (10b) (10c) where I[ ] denotes the indicator function. Here, i j=1 x j represents the number of downloading operations before the (i + 1)-th uploading operation. To increase the transmission time for all tasks so as to reduce the transmission energy consumption for latency-sensitive tasks, we allow the execution of one task to be conducted in parallel with the uploading or downloading of

7 7 (a) A 1 < A 2 < D 2 < D 1, q u = (1, 2), q d = (2, 1) and x = (0, 2). (b) A 1 < A 2 < D 1 < D 2 and q u = q d = (1, 2). Fig. 2. Illustration example of the importance of the choices for q and x, under given (t u, t e, t d ) at K = 2. another task at any time, as illustrated in Fig. 2. C. Energy Consumption Model In this paper, we consider low CPU voltage at the serving node, and adopt the energy consumption model for task execution as in [7], [10]. In particular, the energy consumption for executing task k at the serving node is E e,k µl e,k F 2, where µ is a constant factor determined by the switched capacitance at the MEC server. We consider a narrow band system and study the block fading channel model. Let h k denote the channel power gain for user k, which is assumed to be constant within the duration [A k, D k ]. For simplicity, we consider capacity achieving codes, as in [7], [10]. It can be easily shown that the energy consumption at user k for uploading task ( k to the serving node is E u,k (t u,k ) t u,k Lu,k h k g 2 t u,k ), and the energy consumption at the serving

8 8 ( node for transmitting the computation result of task k to user k is E d,k (t d,k ) t d,k Ld,k h k g 2 t d,k ), where g(x) n 0 ( 2 x B 1 ), and B and n 0 denote the bandwidth (in Hz) and the power (in Watt) of the complex additive white Gaussian channel noise, respectively. Thus, the weighted sum energy consumption for serving task k is E k (t u,k, t d,k ) = E u,k (t u,k ) + β(e e,k + E d,k (t d,k )), where 0 < β 1 is the corresponding weight factor. Note that 0 < β < 1 means imposing a higher cost on the energy consumption for user devices due to their limited battery power. Therefore, the weighted sum energy consumption for serving all K tasks is given by: E(t) k K E k (t u,k, t d,k ). (11) Note that E(t) is a convex function of t. III. PROBLEM FORMULATION AND OPTIMAL SOLUTION In this section, we first formulate the energy minimization problem. Then, we develop an algorithm to obtain an optimal solution by exploiting structural properties of the problem. A. Problem Formulation In this part, we would like to minimize the weighted sum energy consumption by optimizing the operation sequences q o, o O and x, operation starting times s and transmission durations t. Problem 1 (Energy Minimization): E min q,x,s,t E(t) s.t. (1), (3), (4), (5), (6), (7), (8), (9), (10), where E(t) is given by (11). Problem 1 is a mixed discrete-continuous optimization problem with two types of variables, i.e., the operation sequences (discrete variables) as well as the uploading and downloading durations and starting times for uploading, executing and downloading (continuous variables). Note that the set of possible choices for the discrete variables, {(q, x) : (6), (8), (9)}, has cardinality (K!) 3( 2K 1 K 1 ), which is prohibitively large for large K. Thus, Problem 1 is very challenging, and is in general NP-hard.

9 9 B. Optimal Solution In this part, we develop an algorithm to obtain an optimal solution of Problem 1. 1) Structural Properties: First, we analyze structural properties of Problem 1. For all q o Q and o O, let idx(q o, i) denote the index of element i in q o, i.e., q o,idx(qo,i) = i. 3 For all feasible (q, x) of Problem 1, we have the following results. Lemma 1 (Structural Properties of Problem 1): (i) For all k K, idx(q u,k) 1 x i < idx(q d, k). (ii) For all k, k K with idx(q d, k) idx(q u,k ) 1 x i, idx(q e, k) < idx(q e, k ). (iii) For all k, k K with D k < A k, idx(q u, k) < idx(q u, k ), idx(q d, k) < idx(q d, k ) and idx(q d, k) idx(qu,k ) 1 x i. Proof 1: Please refer to Appendix. A. Property (i) indicates that the uploading of task k should be conducted before its downloading. Property (ii) indicates that if the downloading of task k is conducted before the uploading of task k, the executing of task k should be conducted before the executing of task k. Property (iii) indicates that if the deadline of task k is before the arrival time of task k, the uploading of task k should be conducted before the uploading of task k, the downloading of task k should be conducted before the downloading of task k, and the downloading of task k should be conducted before the uploading of task k. Lemma 1 offers important design insights for choosing appropriate operation sequences. Later, we shall see that Lemma 1 also facilitates the process of solving Problem 1. 2) Algorithm: It can be seen that the structural properties for discrete variables (q, x) in Lemma 1 reflect the constraints on continuous variables (s, t) in (3), (4) and (5) to certain extent. Based on Lemma 1, we first propose an equivalent formulation of Problem 1 to facilitate the optimization. Problem 2 (Operation Sequences): E = min Eseq(q, x), (q,x) S where S {(q, x) : (6), (8), (9), Properties (i), (ii) and (iii)}. Let (q, x ) denote an optimal solution. E seq(q, x) is given by the following problem. 3 We assume 0 xi = 0.

10 10 Algorithm 1 : Algorithm for Obtaining An Optimal Solution Output: (q, x, s, t ). 1: Set E = 2: for each (q, x) S do 3: Obtain (s (q, x), t (q, x)) and E seq(q, x) by solving Problem 3 4: if E seq(q, x) E then 5: Set E = E seq(q, x), (q, x, s, t ) = (q, x, s (q, x), t (q, x)) 6: end if 7: end for Problem 3 (Starting Times and Transmission Durations): For any (q, x) S, E seq(q, x) = min s,t E(t), s.t. (1), (3), (4), (5), (7), (10). If the problem is infeasible, we have Eseq(q, x) = ; Otherwise, we have Eseq(q, x) <, and denote an optimal solution by (s (q, x), t (q, x)). Due to the equivalence between Problem 1 and Problems 2-3, we know that (q, x, s (q, x ), t (q, x )) is an optimal solution of Problem 1. This equivalent formulation separates the discrete and continuous variables, and enables solving a mixed discrete-continuous optimization problem (i.e., Problem 1) by solving a discrete optimization problem (i.e., Problem 2) and a continuous optimization problem for each (q, x) S (i.e., Problem 3). Note that for any given (q, x) S, Problem 3 is a convex optimization problem with 5K variables, and thus can be solved using standard convex optimization techniques. Problem 2 is a discrete optimization problem with 4K variables, and can be solved by exhaustive search over set S. Note that the structural properties in Lemma 1 enable a great reduction of the search space for the values of the discrete variables of Problem 2 (from {(q, x) : (6), (8), (9)} to S). The details for solving Problem 1 based on the equivalent formulation in Problem 2 and Problem 3 are summarized in Algorithm 1. IV. LOW-COMPLEXITY SUBOPTIMAL SOLUTION Although the complexity for obtaining an optimal solution of Problem 1 has been reduced based on Lemma 1, the computation complexity of Algorithm 1 may not be acceptable when K is large. In this section, we propose a low-complexity algorithm to obtain a suboptimal solution of Problem 1.

11 11 First, using several optimization techniques, we transform the mixed discrete-continuous problem (i.e., Problem 1) to a DC problem. Denote Q o (Q o,i,j ) i,j K, o O, Q (Q o ) o O, X (X i,j ) i,j K, Z (Z i,j ) i,j K, α o (α o,i,j ) i,j K, o O, α (α o ) o O, φ o (φ o,i,j ) i,j K, o O, φ (φ o ) o O, λ (λ i,n,j ) j K,i,n K and γ (γ i,m,j ) m,j K,i K. 4 Problem 4 (DC Problem of Problem 1): Ẽ min 0 Q,X,Z 1, s,t,α,φ,λ,γ 0 E(t) s.t. (1), (3), (4), (5), P (Q, X, Z) 0, (12) Q o,i,j = Q o,j,i = 1, o O, j K (13) i K i K X i,j = K, (14) i,j K Z i,j = K 1, i,j K j K X m,n (i 1) K (1 Z i,j ), i, j K (16) m=1 n=1 j K α o,i,j j K (15) φ o,i+1,j, o O, i K (17) φ o,i,j min{ sq o,i,j, s o,j }, o O, i, j K s o,j (1 Q o,i,j ) s φ o,i,j, o O, i, j K α o,i,j min{( s + t)q o,i,j, s o,j + t o,j }, o O, i, j K s o,j + t o,j (1 Q o,i,j )( s + t) α o,i,j, o O, i, j K (18a) (18b) (18c) (18d) 4 In this paper, and represent componentwise inequities.

12 12 λ i,m,j min {( s + t)q d,m,j, ( s + t)z m+1,i, s d,j + t d,j }, j K, i, m K s d,j + t d,j (2 Q d,m,j Z m+1,i ) ( s + t) λ i,m,j, j K, i, m K γ i,n,j min{ sq d,n,j, sz n,i, s d,j }, n, j K, i K s d,j (2 Q d,n,j Z n,i ) s γ i,n,j, n, j K, i K (19a) (19b) (19c) (19d) λ i,m,j φ u,i+1,j, i K j K m K j K α u,i+1,j γ i,n,j, i K j K j K n K α u,1,j φ d,1,j, j K j K (20a) (20b) (20c) where s max k K D k, t max k K D k min k K A k, P (Q, X, Z) is given by P (Q, X, Z) i,j K (X i,j (1 X i,j ) + Z i,j (1 Z i,j )) + o O i,j K Let (Q, X, Z, s, t, α, φ, λ, γ ) denote an optimal solution of Problem 4. Q o,i,j (1 Q o,i,j ). (21) Note that the constraints in (12), (13) and (14) correspond to the constraints in (6), (8) and (9), the constraints in (12), (13), (17) and (18) correspond to the constraints in (7), the constraints in (12), (13), (14), (15), (16), (18), (19) and (20) correspond to the constraints in (10). The relationship between Problem 1 and Problem 4 is shown below. Theorem 1 (Equivalence between Problem 1 and Problem 4): Problem 4 is equivalent to Problem 1, i.e., q o,i = K j=1 jq o,i,j, o O, i K, x i = K j=1 X i,j, i K, s o,k = s o,k, o O, k K, t u,k = t u,k, k K, t d,k = t d,k, k K, and E = Ẽ. Proof 2: Firstly, we equivalently convert sequences q and x satisfying (6), (8) and (9) to binary variables Q and X satisfying (13), (14) and X i,j, Q o,i,j {0, 1}, o O, i, j K. (22) The equivalence can be seen from Q o,i,j = I[q o,i = j], q o,i = j K ji[q o,i,j = 1], X i,j = I[j x i ] and x i = j K X i,j. Then, we introduce new variables Z i,j I[ j m=1 K n=1 X m,n = (i

13 13 1)], i, j K and impose (15) and (16). Thus, we can equivalently convert (7) and (10) to m K Z m+1,i Q u,i+1,j s u,j, i K, (23) Q d,m,j (s d,j + t d,j ) j K j K Z n,i Q d,n,j s d,j, i K. (24) Q u,i+1,j (s u,j + t u,j ) j K n K j K Q o,i+1,j s o,j, o O, i K, (25) j K Q o,i,j (s o,j + t o,j ) j K j K Q u,1,j (s u,j + t u,j ) j K Q d,1,j s d,j. (26) Secondly, we adopt the big-m formulation to decompose the product terms Q o,i,j (s o,j + t o,j ), Q o,i+1,j s o,j, Z m+1,i Q d,m,j (s d,j +t d,j ), Z n,i Q d,n,j s d,j in (23), (24), (25) and (26) by introducing new variables α, φ, λ, γ and imposing the linear constraints in (18) and (19). Finally, we equivalently convert the discrete constraints in (22) to continuous constraints in (12) and 0 Q, X, Z 1. By noting that x a (a 0) can be rewritten as x a and a x, and x min{a 1, a 2,..., a n } can be rewritten as x a i, i = 1, 2,..., n, the constraints in (16), (18) and (19) can be viewed as linear constraints. In addition, note that P (Q, X, Z) given by (21) is concave, the other constraints are linear, and E(t) is convex. Thus, Problem 4 is a DC problem, which is non-convex. A classical goal of solving a non-convex problem is to obtain a stationary point. Next, we obtain a stationary point of Problem 4 using Penalty CCP [9]. Specifically, at the (l + 1)-th iteration, we solve the following convex approximation problem, which is obtained via linearizing the concave portion in P (Q, X, Z) at the solution obtained at the l-th iteration, and relaxing the constraint in (12) by introducing a slack variable y. Problem 5 (Convex Approximation Problem at (l + 1)-th iteration): arg min 0 Q,X,Z 1, s,t,α,φ,λ,γ 0,y 0 k K E(t) + ρ (l+1) y s.t. (1), (3), (4), (5), (13), (14), (15), (16), (17), (18), (19), (20) ˆP ( Q, X, Z; Q (l), X (l), Z (l)) y, where ρ (l+1) = min{θρ (l), ρ} for some θ > 1 and ρ > 0, and ˆP ( Q, X, Z; Q (l), X (l), Z (l)) is

14 14 Algorithm 2 : Algorithm for Obtaining A Low-complexity Suboptimal Solution ) 1: Initialization: choose a feasible initial point (Q (0), X (0), Z (0), s (0), t (0), α (0), φ (0), λ (0), γ (0) of Problem 4, and set ρ (0) > 0 and l = 0. 2: repeat ( 3: Obtain Q (l+1), X (l+1), Z (l+1), s (l+1), t (l+1), α (l+1), ) φ (l+1), λ (l+1), γ (l+1) by solving Problem 5 using standard convex optimization techniques. 4: Set ρ (l+1) = min{θρ (l), ρ} and l = l + 1 5: until convergence criteria is met given by ˆP ( Q, X, Z; Q (l), X (l), Z (l)) (( 1 2Z (l) i,j i,j K i,j K o O ) ) Z i,j + Z (l) 2 i,j + i,j K (( + ((( ) )) 1 2Q (l) o,i,j Q o,i,j + Q (l) 2 o,i,j 1 2X (l) i,j ) ) X i,j + X (l) 2 i,j ) Let (Q (l+1), X (l+1), Z (l+1), s (l+1), t (l+1), α (l+1), φ (l+1), λ (l+1), γ (l+1), y (l+1) denote an optimal solution of Problem 5 at the (l + 1)-th iteration. The details of the algorithm are summarized in Algorithm 2. By [9], we know that the sequence )} {(Q (l), X (l), Z (l), s (l), t (l), α (l), φ (l), λ (l), γ (l) generated by Algorithm 2 is convergent, )} and the limit point of {(Q (l), X (l), Z (l), s (l), t (l), α (l), φ (l), λ (l), γ (l) is a stationary point of Problem 4. We can run Algorithm 2 multiple times, each with a random initial feasible point of Problem 4, and select the stationary point with the minimum weighted sum energy, denoted by ( Q, X, Z, s, t ). Then, we can construct a suboptimal solution of Problem 1, denoted by ( q, x, s, t ), where q o,i = K j=1 jq o,i,j, o O, i K and x i = K j=1 X i,j, i K. V. NUMERICAL RESULTS In this section, we compare the proposed optimal and suboptimal solutions with two optimized baselines schemes in three different cases of task arrival times and deadlines as illustrated in Fig. 3 using numerical results. Both baseline schemes allow parallel transmissions and executions and consider the optimization of starting times and transmission durations but for two simple choices of operation sequences, i.e., adopt the solutions of Problem 2 for two typical (q, x) S. Our primary goal is to demonstrate the key impact of the operation sequences in designing efficient MEC systems.

15 15 Fig. 3. Illustration of the three cases of task arrival times and deadlines. Let v A Q and v D Q denote the orders of the arrival times and deadlines of the K tasks, where v A,i and v D,i represent the indices of the tasks corresponding to the i-th smallest arrival time and deadline, respectively. Both baseline schemes choose q u = q e = v A, which seems quite reasonable. In addition, Baseline 1 chooses x with x i = 0, i K, x K = K and q d = v D, and Baseline 2 chooses x with x i = 1, i K and q d = v A. That is, in Baseline 1, the uploading operations of all K tasks are completed before the downloading operation of any task, and in Baseline 2, the uploading and downloading operations of a task are completed before those of another task. Note that the two choices for q d and x are also typical. The considered three cases have different arrival times and deadlines. As illustrated in Fig. 3, in Case 1, the order of arrival times is the reverse of that of deadlines, i.e., v A,i = v D,K+1 i, i K; in Case 2, the order of arrival times is the same as that of deadlines, i.e., v A,i = v D,i, i K; in Case 3, the order of arrival times and the order of deadlines can be treated as a combination of those in Case 1 and Case 2, i.e., v A,i = v D,i I[i is odd] + v D,2 I[i is even], i K, where denotes the floor K 2 +2 i function. In the simulation, we consider the following settings. For those three cases, we set L u,k = k 10 4 bits, L d,k = 1.5k 10 4 bits, L e,k = 1 K 105 CPU-cycles, h k = 10 3, k K, β = 0.1, µ = 10 29, F = , B = 10 6 Hz and n 0 = k B BT 0, where k B = Joule/Kelvin is the Boltzmann constant and T 0 = 300 Kelvin is the temperature. For ease of illustration, let

16 16 Energy consumption(j) Optimal Suboptimal Baseline 1 Baseline Number of users K (a) Energy consumption versus K. Computation Time (s) Optimal Suboptimal Baseline 1 Baseline Number of users K (b) Computation time versus K. Fig. 4. Comparisions between proposed solutions and baseline schemes in Case 1, where A i = 15 K, D i = 30 K and D 1 A 1 = K 15 K. Energy consumption(j) Optimal Suboptimal Baseline 1 Baseline Number of users K (a) Energy consumption versus K. Computation Time (s) Optimal Suboptimal Baseline 1 Baseline Number of users K (b) Computation time versus K. Fig. 5. Comparisions between proposed solutions and baseline schemes in Case 2, where A i = 15 K, D i = 45 K and D 1 A 1 = 6K + 30 K. A i A i+1 A i, D i D i+1 D i, i K, and we consider time in ms. Fig. 4, Fig. 5 and Fig. 6 illustrate the weighted sum energy consumption and computation time (reflecting computation complexity) of the proposed solutions and baseline schemes versus the number of users K in the three cases, respectively. Note that since the computation complexity for obtaining an optimal solution is prohibitively high when K is large, the proposed optimal solution is only evaluated at K = 2, 3. In addition, note that in Case 3, Baseline 1 is not feasible when K 5. From Fig. 4(a), Fig. 5(a) and Fig. 6(a), we can observe that the weighted sum

17 17 Energy consumption(j) Optimal Suboptimal Baseline 1 Baseline Number of users K (a) Energy consumption versus K. Computation Time (s) Optimal Suboptimal Baseline 1 Baseline Number of users K (b) Computation time versus K. Fig. 6. Comparisions between proposed solutions and baseline schemes in Case 3, where A i = 12I[i is odd] + 3I[i is even], D i + D i+1 = 25I[i is odd] 11I[i is even], D 1 A 1 = 25 and D 2 A 2 = K 2. energy consumption of the proposed suboptimal solution is close to that of the optimal solution when K = 2, 3 in the three cases. In addition, the proposed suboptimal solution achieves smaller weighted sum energy consumption than Baseline 2 in Case 1, Baseline 1 in Case 2, and both baseline schemes in Case 3 (e.g., 97% smaller than Baseline 1 and 66% smaller than Baseline 2 both at K = 4). This observation demonstrates that the typical choices for operation sequences may perform well in some special cases of task arrival times and deadlines, but cannot guarantee universally good performance in the general scenario. From Fig. 4(b), Fig. 5(b) and Fig. 6(b), we see that the computation complexity of the optimal solution is not acceptable when K 4, and the computation complexity of the proposed suboptimal solution is quite close to those of both baseline schemes. These numerical results demonstrate the applicability and effectiveness of the proposed suboptimal solution. VI. CONCLUSION In this paper, we considered a computation task model and a computation offloading model, which are more practical than the existing ones. Then, we formulated the weighted sum energy consumption minimization problem to optimize the task operation sequences and starting times for uploading, executing and downloading as well as uploading and downloading time durations. The problem is a challenging mixed discrete-continuous optimization problem. By carefully using several optimization techniques, we developed an algorithm to obtain an optimal solution and a

18 18 low-complexity algorithm to obtain a suboptimal solution. Both analytical and numerical results demonstrate the importance of the optimization of operation sequences in practical multi-user MEC systems. A. Proof of Lemma 1 APPENDIX First, we prove Property (i) by contradiction. Suppose that there exists k K satisfying idx(qu,k ) 1 x i idx(q d, k ). By (7) and idx(q u,k ) 1 x i idx(q d, k ), we have: s d,k + t d,k = s + t d,qd,idx(qd,k ) d,q s d,idx(qd,k ) d,q. (27) idx(qu,k ) 1 d, x i By (10), we have: By (27) and (28), we have: s d,qd, s idx(qu,k ) 1 u,qu,idx(qu,k ) = s u,k. (28) x i s d,k s u,k. (29) In addition, by (2) and (3), we have: s u,k s e,k < s d,k. (30) It is clear that (30) contradicts (29). Thus, by contradiction, we can prove Property (i). Next, we prove Property (ii) by contradiction. Suppose that there exist k, k K satisfying idx(q d, k) idx(q u,k ) 1 x i and idx(q e, k) idx(q e, k ). By (3), (7), (10) and idx(q d, k) idx(qu,k ) 1 x i, we have: s e,k < s d,k + t d,k = s + t d,qd,idx(qd,k) d,q s d,idx(qd,k) d,q s idx(qu,k ) 1 u,k s e,k. (31) d, x i In addition, by (7) and idx(q e, k) idx(q e, k ), we have: s e,k s e,k. (32) It is clear that (32) contradicts (31). Thus, by contradiction, we can prove Property (ii). Finally, we prove Property (iii) by contradiction. For all k, k K, by (2), (3) and (7), we

19 19 have: A k s u,k s e,k < s d,k D k, (33) A k s u,k s e,k < s d,k D k. (34) 1) Suppose that there exist k, k K with D k < A k satisfying idx(q u, k) idx(q u, k ). By (7) and idx(q u, k) idx(q u, k ), we have: s u,k s u,k (35) By (33), (34) and D k < A k, we have: s u,k < s u,k (36) It is clear that (35) contradicts (36). 2) Suppose that there exist k, k K with D k < A k satisfying idx(q d, k) idx(q d, k ). By (7) and idx(q d, k) idx(q d, k ), we have: s d,k s d,k. (37) By (33), (34) and D k < A k, we have: s d,k < s d,k. (38) It is clear that (37) contradicts (38). 3) Suppose that there exist k, k K with D k < A k satisfying idx(q d, k) > idx(q u,k ) 1 x i. By (7), (10) and idx(q d, k) > idx(q u,k ) 1 x i, we have: s u,k = s u,qu,idx(qu,k ) s d,qd, s = s idx(qu,k ) 1 d,qd,idx(qd,k) d,k. (39) x i +1 By (33), (34) and D k < A k, we have: s d,k < s u,k. (40) It is clear that (39) contradicts (40). Thus, by contradiction, we can prove Property (iii). Therefore, we complete the proof of Lemma 1.

20 20 REFERENCES [1] M.-H. Chen, B. Liang, and M. Dong, Joint offloading decision and resource allocation for multi-user multi-task mobile cloud, in IEEE ICC, Kuala Lumpur, Malaysia, [2] Z. Sheng, C. Mahapatra, V. C. Leung, M. Chen, and P. K. Sahu, Energy efficient cooperative computing in mobile wireless sensor networks, IEEE Trans. Cloud Comput., vol. 6, no. 1, pp , [3] K. Wang, K. Yang, and C. S. Magurawalage, Joint energy minimization and resource allocation in c-ran with mobile cloud, IEEE Trans. Cloud Comput., vol. 6, no. 3, pp , [4] L. Yang, H. Zhang, M. Li, J. Guo, and H. Ji, Mobile edge computing empowered energy efficient task offloading in 5g, IEEE Trans. Veh. Technol., vol. 67, no. 7, pp , [5] Y. Sun, S. Zhou, and J. Xu, Emm: Energy-aware mobility management for mobile edge computing in ultra dense networks, IEEE J. Select. Areas Commun., vol. 35, no. 11, pp , [6] Y. Mao, J. Zhang, and K. B. Letaief, Dynamic computation offloading for mobile-edge computing with energy harvesting devices, IEEE J. Select. Areas Commun., vol. 34, no. 12, pp , [7] J. Guo, Z. Song, Y. Cui, Z. Liu, and Y. Ji, Energy-efficient resource allocation for multi-user mobile edge computing, in IEEE GLOBECOM, Singapore, [8] C. You, Y. Zeng, R. Zhang, and K. Huang, Asynchronous mobile-edge computation offloading: energy-efficient resource management, arxiv preprint arxiv: , [9] T. Lipp and S. Boyd, Variations and extension of the convex concave procedure, Optimization and Engineering, vol. 17, no. 2, pp , [10] C. You, K. Huang, H. Chae, and B. Kim, Energy-efficient resource allocation for mobile-edge computation offloading, IEEE Trans. Wireless Commun., vol. 16, no. 3, pp , [11] X. Wang, Y. Cui, Z. Liu, J. Guo, and M. Yang, Optimal resource allocation for multi-user mec with arbitrary task arrival times and deadlines, Technical Report, [Online]. Available:

Energy-Efficient Resource Allocation for Multi-User Mobile Edge Computing

Energy-Efficient Resource Allocation for Multi-User Mobile Edge Computing Energy-Efficient Resource Allocation for Multi-User Mobile Edge Computing Junfeng Guo, Zhaozhe Song, Ying Cui Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai, P. R. China

More information

Energy-Efficient Resource Allocation for Cache-Assisted Mobile Edge Computing

Energy-Efficient Resource Allocation for Cache-Assisted Mobile Edge Computing Energy-Efficient Resource Allocation for Cache-Assisted Mobile Edge Computing Ying Cui, Wen He, Chun Ni, Chengjun Guo Department of Electronic Engineering Shanghai Jiao Tong University, China Zhi Liu Department

More information

Optimal Offloading and Resource Allocation in Mobile-Edge Computing with Inter-user Task Dependency

Optimal Offloading and Resource Allocation in Mobile-Edge Computing with Inter-user Task Dependency Optimal Offloading and Resource Allocation in Mobile-Edge Computing with Inter-user Task Dependency Jia Yan, Suzhi Bi, Member, IEEE, Ying-Jun Angela Zhang, Senior 1 arxiv:1810.11199v1 cs.dc] 26 Oct 2018

More information

THE rapid developments of Internet-of-things (IoT) and

THE rapid developments of Internet-of-things (IoT) and IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, 207 Wireless Powered Cooperation-Assisted Mobile Edge Computing Xiaoyan Hu, Student Member, IEEE, Kai-Kit Wong, Fellow, IEEE, and Kun Yang, Senior Member,

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

Distributed Joint Offloading Decision and Resource Allocation for Multi-User Mobile Edge Computing: A Game Theory Approach

Distributed Joint Offloading Decision and Resource Allocation for Multi-User Mobile Edge Computing: A Game Theory Approach Distributed Joint Offloading Decision and Resource Allocation for Multi-User Mobile Edge Computing: A Game Theory Approach Ning Li, Student Member, IEEE, Jose-Fernan Martinez-Ortega, Gregorio Rubio Abstract-

More information

Computation Offloading Strategy Optimization with Multiple Heterogeneous Servers in Mobile Edge Computing

Computation Offloading Strategy Optimization with Multiple Heterogeneous Servers in Mobile Edge Computing IEEE TRANSACTIONS ON SUSTAINABLE COMPUTING VOL XX NO YY MONTH 019 1 Computation Offloading Strategy Optimization with Multiple Heterogeneous Servers in Mobile Edge Computing Keqin Li Fellow IEEE Abstract

More information

Transmission Schemes for Lifetime Maximization in Wireless Sensor Networks: Uncorrelated Source Observations

Transmission Schemes for Lifetime Maximization in Wireless Sensor Networks: Uncorrelated Source Observations Transmission Schemes for Lifetime Maximization in Wireless Sensor Networks: Uncorrelated Source Observations Xiaolu Zhang, Meixia Tao and Chun Sum Ng Department of Electrical and Computer Engineering National

More information

Analysis and Optimization of Caching and Multicasting for Multi-Quality Videos in Large-Scale Wireless Networks

Analysis and Optimization of Caching and Multicasting for Multi-Quality Videos in Large-Scale Wireless Networks Analysis and Optimization of Caching and Multicasting for Multi-Quality Videos in Large-Scale Wireless Networks Dongdong Jiang, Student Member, IEEE and Ying Cui, Member, IEEE Abstract Efficient dissemination

More information

EP2200 Course Project 2017 Project II - Mobile Computation Offloading

EP2200 Course Project 2017 Project II - Mobile Computation Offloading EP2200 Course Project 2017 Project II - Mobile Computation Offloading 1 Introduction Queuing theory provides us a very useful mathematic tool that can be used to analytically evaluate the performance of

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

On the complexity of maximizing the minimum Shannon capacity in wireless networks by joint channel assignment and power allocation

On the complexity of maximizing the minimum Shannon capacity in wireless networks by joint channel assignment and power allocation On the complexity of maximizing the minimum Shannon capacity in wireless networks by joint channel assignment and power allocation Mikael Fallgren Royal Institute of Technology December, 2009 Abstract

More information

Optimal Power Allocation for Cognitive Radio under Primary User s Outage Loss Constraint

Optimal Power Allocation for Cognitive Radio under Primary User s Outage Loss Constraint This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE ICC 29 proceedings Optimal Power Allocation for Cognitive Radio

More information

BS-assisted Task Offloading for D2D Networks with Presence of User Mobility

BS-assisted Task Offloading for D2D Networks with Presence of User Mobility BS-assisted Task Offloading for D2D Networks with Presence of User Mobility Ghafour Ahani and Di Yuan Department of Information Technology Uppsala University, Sweden Emails:{ghafour.ahani, di.yuan}@it.uu.se

More information

Stochastic Content-Centric Multicast Scheduling for Cache-Enabled Heterogeneous Cellular Networks

Stochastic Content-Centric Multicast Scheduling for Cache-Enabled Heterogeneous Cellular Networks 1 Stochastic Content-Centric Multicast Scheduling for Cache-Enabled Heterogeneous Cellular Networks Bo Zhou, Ying Cui, Member, IEEE, and Meixia Tao, Senior Member, IEEE Abstract Caching at small base stations

More information

Energy-Efficient Real-Time Task Scheduling in Multiprocessor DVS Systems

Energy-Efficient Real-Time Task Scheduling in Multiprocessor DVS Systems Energy-Efficient Real-Time Task Scheduling in Multiprocessor DVS Systems Jian-Jia Chen *, Chuan Yue Yang, Tei-Wei Kuo, and Chi-Sheng Shih Embedded Systems and Wireless Networking Lab. Department of Computer

More information

Short-Packet Communications in Non-Orthogonal Multiple Access Systems

Short-Packet Communications in Non-Orthogonal Multiple Access Systems Short-Packet Communications in Non-Orthogonal Multiple Access Systems Xiaofang Sun, Shihao Yan, Nan Yang, Zhiguo Ding, Chao Shen, and Zhangdui Zhong State Key Lab of Rail Traffic Control and Safety, Beijing

More information

Energy-Efficient Data Transmission with Non-FIFO Packets

Energy-Efficient Data Transmission with Non-FIFO Packets Energy-Efficient Data Transmission with Non-FIFO Packets 1 Qing Zhou, Nan Liu National Mobile Communications Research Laboratory, Southeast University, arxiv:1510.01176v1 [cs.ni] 5 Oct 2015 Nanjing 210096,

More information

NOMA: Principles and Recent Results

NOMA: Principles and Recent Results NOMA: Principles and Recent Results Jinho Choi School of EECS GIST September 2017 (VTC-Fall 2017) 1 / 46 Abstract: Non-orthogonal multiple access (NOMA) becomes a key technology in 5G as it can improve

More information

NOMA: An Information Theoretic Perspective

NOMA: An Information Theoretic Perspective NOMA: An Information Theoretic Perspective Peng Xu, Zhiguo Ding, Member, IEEE, Xuchu Dai and H. Vincent Poor, Fellow, IEEE arxiv:54.775v2 cs.it] 2 May 25 Abstract In this letter, the performance of non-orthogonal

More information

Call Completion Probability in Heterogeneous Networks with Energy Harvesting Base Stations

Call Completion Probability in Heterogeneous Networks with Energy Harvesting Base Stations Call Completion Probability in Heterogeneous Networks with Energy Harvesting Base Stations Craig Wang, Salman Durrani, Jing Guo and Xiangyun (Sean) Zhou Research School of Engineering, The Australian National

More information

Optimal Power Allocation With Statistical QoS Provisioning for D2D and Cellular Communications Over Underlaying Wireless Networks

Optimal Power Allocation With Statistical QoS Provisioning for D2D and Cellular Communications Over Underlaying Wireless Networks IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, ACCEPTED AND TO APPEAR IN 05 Optimal Power Allocation With Statistical QoS Provisioning for DD and Cellular Communications Over Underlaying Wireless Networks

More information

Traversing Virtual Network Functions from the Edge to the Core: An End-to-End Performance Analysis

Traversing Virtual Network Functions from the Edge to the Core: An End-to-End Performance Analysis Traversing Virtual Network Functions from the Edge to the Core: An End-to-End Performance Analysis Emmanouil Fountoulakis, Qi Liao, Manuel Stein, Nikolaos Pappas Department of Science and Technology, Linköping

More information

Cellular-Enabled UAV Communication: Trajectory Optimization Under Connectivity Constraint

Cellular-Enabled UAV Communication: Trajectory Optimization Under Connectivity Constraint Cellular-Enabled UAV Communication: Trajectory Optimization Under Connectivity Constraint Shuowen Zhang, Yong Zeng, and Rui Zhang ECE Department, National University of Singapore. Email: {elezhsh,elezeng,elezhang}@nus.edu.sg

More information

Joint and Competitive Caching Designs in Large-Scale Multi-Tier Wireless Multicasting Networks

Joint and Competitive Caching Designs in Large-Scale Multi-Tier Wireless Multicasting Networks Joint and Competitive Caching Designs in Large-Scale Multi-Tier Wireless Multicasting Networks Zitian Wang, Zhehan Cao, Ying Cui Shanghai Jiao Tong University, China Yang Yang Intel Deutschland GmbH, Germany

More information

Lecture 13. Real-Time Scheduling. Daniel Kästner AbsInt GmbH 2013

Lecture 13. Real-Time Scheduling. Daniel Kästner AbsInt GmbH 2013 Lecture 3 Real-Time Scheduling Daniel Kästner AbsInt GmbH 203 Model-based Software Development 2 SCADE Suite Application Model in SCADE (data flow + SSM) System Model (tasks, interrupts, buses, ) SymTA/S

More information

Exploiting Non-Causal CPU-State Information for Energy-Efficient Mobile Cooperative Computing

Exploiting Non-Causal CPU-State Information for Energy-Efficient Mobile Cooperative Computing 1 Exploiting Non-Causal CPU-State Information for Energy-Efficient Mobile Cooperative Computing Changsheng You and Kaibin Huang Abstract arxiv:1704.04595v4 [cs.it] 4 Sep 2017 Scavenging the idling computation

More information

Performance Analysis of a Threshold-Based Relay Selection Algorithm in Wireless Networks

Performance Analysis of a Threshold-Based Relay Selection Algorithm in Wireless Networks Communications and Networ, 2010, 2, 87-92 doi:10.4236/cn.2010.22014 Published Online May 2010 (http://www.scirp.org/journal/cn Performance Analysis of a Threshold-Based Relay Selection Algorithm in Wireless

More information

Game Theoretic Approach to Power Control in Cellular CDMA

Game Theoretic Approach to Power Control in Cellular CDMA Game Theoretic Approach to Power Control in Cellular CDMA Sarma Gunturi Texas Instruments(India) Bangalore - 56 7, INDIA Email : gssarma@ticom Fernando Paganini Electrical Engineering Department University

More information

Resource Allocation for Wireless Fading Relay Channels: Max-Min Solution 1 2

Resource Allocation for Wireless Fading Relay Channels: Max-Min Solution 1 2 Submitted to IEEE Trans. Inform. Theory, Special Issue on Models, Theory and odes for elaying and ooperation in ommunication Networs, Aug. 2006, revised Jan. 2007 esource Allocation for Wireless Fading

More information

Optimal matching in wireless sensor networks

Optimal matching in wireless sensor networks Optimal matching in wireless sensor networks A. Roumy, D. Gesbert INRIA-IRISA, Rennes, France. Institute Eurecom, Sophia Antipolis, France. Abstract We investigate the design of a wireless sensor network

More information

Optimal power allocation on discrete energy harvesting model

Optimal power allocation on discrete energy harvesting model Wang et al. EURASIP Journal on Wireless Communications and Networking (205) 205:48 DOI 0.86/s3638-05-028-x RESEARCH Open Access Optimal power allocation on discrete energy harvesting model Xiaolei Wang,JieGong

More information

Transmitter-Receiver Cooperative Sensing in MIMO Cognitive Network with Limited Feedback

Transmitter-Receiver Cooperative Sensing in MIMO Cognitive Network with Limited Feedback IEEE INFOCOM Workshop On Cognitive & Cooperative Networks Transmitter-Receiver Cooperative Sensing in MIMO Cognitive Network with Limited Feedback Chao Wang, Zhaoyang Zhang, Xiaoming Chen, Yuen Chau. Dept.of

More information

On the Power Allocation for Hybrid DF and CF Protocol with Auxiliary Parameter in Fading Relay Channels

On the Power Allocation for Hybrid DF and CF Protocol with Auxiliary Parameter in Fading Relay Channels On the Power Allocation for Hybrid DF and CF Protocol with Auxiliary Parameter in Fading Relay Channels arxiv:4764v [csit] 4 Dec 4 Zhengchuan Chen, Pingyi Fan, Dapeng Wu and Liquan Shen Tsinghua National

More information

Morning Session Capacity-based Power Control. Department of Electrical and Computer Engineering University of Maryland

Morning Session Capacity-based Power Control. Department of Electrical and Computer Engineering University of Maryland Morning Session Capacity-based Power Control Şennur Ulukuş Department of Electrical and Computer Engineering University of Maryland So Far, We Learned... Power control with SIR-based QoS guarantees Suitable

More information

Wireless Networks for Mobile Edge Computing: Spatial Modeling and Latency Analysis

Wireless Networks for Mobile Edge Computing: Spatial Modeling and Latency Analysis Wireless Networks for Mobile Edge Computing: Spatial Modeling and Latency Analysis Seung-Woo Ko, Kaifeng Han, and Kaibin Huang arxiv:709.0702v2 [cs.it] 9 Jan 208 Abstract Next-generation wireless networks

More information

Energy-Efficient Resource Allocation for

Energy-Efficient Resource Allocation for Energy-Efficient Resource Allocation for 1 Wireless Powered Communication Networks Qingqing Wu, Student Member, IEEE, Meixia Tao, Senior Member, IEEE, arxiv:1511.05539v1 [cs.it] 17 Nov 2015 Derrick Wing

More information

EMM: Energy-Aware Mobility Management for Mobile Edge Computing in Ultra Dense Networks

EMM: Energy-Aware Mobility Management for Mobile Edge Computing in Ultra Dense Networks 1 EMM: Energy-Aware Mobility Management for Mobile Edge Computing in Ultra Dense Networks Yuxuan Sun, Sheng Zhou, Member, IEEE, and Jie Xu, Member, IEEE arxiv:1709.02582v1 [cs.it] 8 Sep 2017 Abstract Merging

More information

TRANSMISSION STRATEGIES FOR SINGLE-DESTINATION WIRELESS NETWORKS

TRANSMISSION STRATEGIES FOR SINGLE-DESTINATION WIRELESS NETWORKS The 20 Military Communications Conference - Track - Waveforms and Signal Processing TRANSMISSION STRATEGIES FOR SINGLE-DESTINATION WIRELESS NETWORKS Gam D. Nguyen, Jeffrey E. Wieselthier 2, Sastry Kompella,

More information

IN this paper, we show that the scalar Gaussian multiple-access

IN this paper, we show that the scalar Gaussian multiple-access 768 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 50, NO. 5, MAY 2004 On the Duality of Gaussian Multiple-Access and Broadcast Channels Nihar Jindal, Student Member, IEEE, Sriram Vishwanath, and Andrea

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

Robust Network Codes for Unicast Connections: A Case Study

Robust Network Codes for Unicast Connections: A Case Study Robust Network Codes for Unicast Connections: A Case Study Salim Y. El Rouayheb, Alex Sprintson, and Costas Georghiades Department of Electrical and Computer Engineering Texas A&M University College Station,

More information

Energy-Sensitive Cooperative Spectrum Sharing: A Contract Design Approach

Energy-Sensitive Cooperative Spectrum Sharing: A Contract Design Approach Energy-Sensitive Cooperative Spectrum Sharing: A Contract Design Approach Zilong Zhou, Xinxin Feng, Xiaoying Gan Dept. of Electronic Engineering, Shanghai Jiao Tong University, P.R. China Email: {zhouzilong,

More information

Distributed Power Control for Time Varying Wireless Networks: Optimality and Convergence

Distributed Power Control for Time Varying Wireless Networks: Optimality and Convergence Distributed Power Control for Time Varying Wireless Networks: Optimality and Convergence Tim Holliday, Nick Bambos, Peter Glynn, Andrea Goldsmith Stanford University Abstract This paper presents a new

More information

QUANTIZATION FOR DISTRIBUTED ESTIMATION IN LARGE SCALE SENSOR NETWORKS

QUANTIZATION FOR DISTRIBUTED ESTIMATION IN LARGE SCALE SENSOR NETWORKS QUANTIZATION FOR DISTRIBUTED ESTIMATION IN LARGE SCALE SENSOR NETWORKS Parvathinathan Venkitasubramaniam, Gökhan Mergen, Lang Tong and Ananthram Swami ABSTRACT We study the problem of quantization for

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

A Half-Duplex Cooperative Scheme with Partial Decode-Forward Relaying

A Half-Duplex Cooperative Scheme with Partial Decode-Forward Relaying A Half-Duplex Cooperative Scheme with Partial Decode-Forward Relaying Ahmad Abu Al Haija, and Mai Vu, Department of Electrical and Computer Engineering McGill University Montreal, QC H3A A7 Emails: ahmadabualhaija@mailmcgillca,

More information

Characterization of Convex and Concave Resource Allocation Problems in Interference Coupled Wireless Systems

Characterization of Convex and Concave Resource Allocation Problems in Interference Coupled Wireless Systems 2382 IEEE TRANSACTIONS ON SIGNAL PROCESSING, VOL 59, NO 5, MAY 2011 Characterization of Convex and Concave Resource Allocation Problems in Interference Coupled Wireless Systems Holger Boche, Fellow, IEEE,

More information

certain class of distributions, any SFQ can be expressed as a set of thresholds on the sufficient statistic. For distributions

certain class of distributions, any SFQ can be expressed as a set of thresholds on the sufficient statistic. For distributions Score-Function Quantization for Distributed Estimation Parvathinathan Venkitasubramaniam and Lang Tong School of Electrical and Computer Engineering Cornell University Ithaca, NY 4853 Email: {pv45, lt35}@cornell.edu

More information

Dynamic Spectrum Leasing with Two Sellers

Dynamic Spectrum Leasing with Two Sellers Dynamic Spectrum Leasing with Two Sellers Rongfei Fan, Member, IEEE, Wen Chen, Hai Jiang, Senior Member, IEEE, Jianping An, Member, IEEE, Kai Yang, Member, IEEE, and Chengwen Xing, Member, IEEE arxiv:62.05702v

More information

Energy-Efficient Admission of Delay-Sensitive Tasks for Mobile Edge Computing

Energy-Efficient Admission of Delay-Sensitive Tasks for Mobile Edge Computing Energy-Efficient Admission of Delay-Sensitive Tasks for Mobile Edge Computing Xinchen Lyu, Hui Tian, Wei Ni, Yan Zhang, Ping Zhang, and Ren Ping Liu Abstract Task admission is critical to delay-sensitive

More information

Spectral and Energy Efficient Wireless Powered IoT Networks: NOMA or TDMA?

Spectral and Energy Efficient Wireless Powered IoT Networks: NOMA or TDMA? 1 Spectral and Energy Efficient Wireless Powered IoT Networs: NOMA or TDMA? Qingqing Wu, Wen Chen, Derric Wing wan Ng, and Robert Schober Abstract Wireless powered communication networs WPCNs, where multiple

More information

Power Control in Multi-Carrier CDMA Systems

Power Control in Multi-Carrier CDMA Systems A Game-Theoretic Approach to Energy-Efficient ower Control in Multi-Carrier CDMA Systems Farhad Meshkati, Student Member, IEEE, Mung Chiang, Member, IEEE, H. Vincent oor, Fellow, IEEE, and Stuart C. Schwartz,

More information

Continuous-Model Communication Complexity with Application in Distributed Resource Allocation in Wireless Ad hoc Networks

Continuous-Model Communication Complexity with Application in Distributed Resource Allocation in Wireless Ad hoc Networks Continuous-Model Communication Complexity with Application in Distributed Resource Allocation in Wireless Ad hoc Networks Husheng Li 1 and Huaiyu Dai 2 1 Department of Electrical Engineering and Computer

More information

Cooperative Diamond Channel With Energy Harvesting Nodes Berk Gurakan, Student Member, IEEE, and Sennur Ulukus, Fellow, IEEE

Cooperative Diamond Channel With Energy Harvesting Nodes Berk Gurakan, Student Member, IEEE, and Sennur Ulukus, Fellow, IEEE 1604 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 34, NO. 5, MAY 016 Cooperative Diamond Channel With Energy Harvesting Nodes Berk Gurakan, Student Member, IEEE, and Sennur Ulukus, Fellow, IEEE

More information

EDF Feasibility and Hardware Accelerators

EDF Feasibility and Hardware Accelerators EDF Feasibility and Hardware Accelerators Andrew Morton University of Waterloo, Waterloo, Canada, arrmorton@uwaterloo.ca Wayne M. Loucks University of Waterloo, Waterloo, Canada, wmloucks@pads.uwaterloo.ca

More information

4888 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 7, JULY 2016

4888 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 7, JULY 2016 4888 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 7, JULY 2016 Online Power Control Optimization for Wireless Transmission With Energy Harvesting and Storage Fatemeh Amirnavaei, Student Member,

More information

Delay QoS Provisioning and Optimal Resource Allocation for Wireless Networks

Delay QoS Provisioning and Optimal Resource Allocation for Wireless Networks Syracuse University SURFACE Dissertations - ALL SURFACE June 2017 Delay QoS Provisioning and Optimal Resource Allocation for Wireless Networks Yi Li Syracuse University Follow this and additional works

More information

Power Allocation and Coverage for a Relay-Assisted Downlink with Voice Users

Power Allocation and Coverage for a Relay-Assisted Downlink with Voice Users Power Allocation and Coverage for a Relay-Assisted Downlink with Voice Users Junjik Bae, Randall Berry, and Michael L. Honig Department of Electrical Engineering and Computer Science Northwestern University,

More information

Opportunistic Spectrum Access for Energy-Constrained Cognitive Radios

Opportunistic Spectrum Access for Energy-Constrained Cognitive Radios 1206 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 8, NO. 3, MARCH 2009 Opportunistic Spectrum Access for Energy-Constrained Cognitive Radios Anh Tuan Hoang, Ying-Chang Liang, David Tung Chong Wong,

More information

Green Distributed Storage Using Energy Harvesting Nodes

Green Distributed Storage Using Energy Harvesting Nodes 1 Green Distributed Storage Using Energy Harvesting Nodes Abdelrahman M. Ibrahim, Student Member, IEEE, Ahmed A. Zewail, Student Member, IEEE, and Aylin Yener, Fellow, IEEE Abstract We consider a distributed

More information

Optimum Transmission through the Gaussian Multiple Access Channel

Optimum Transmission through the Gaussian Multiple Access Channel Optimum Transmission through the Gaussian Multiple Access Channel Daniel Calabuig Institute of Telecommunications and Multimedia Applications Universidad Politécnica de Valencia Valencia, Spain Email:

More information

Front Inform Technol Electron Eng

Front Inform Technol Electron Eng Interference coordination in full-duplex HetNet with large-scale antenna arrays Zhao-yang ZHANG, Wei LYU Zhejiang University Key words: Massive MIMO; Full-duplex; Small cell; Wireless backhaul; Distributed

More information

Energy Cooperation and Traffic Management in Cellular Networks with Renewable Energy

Energy Cooperation and Traffic Management in Cellular Networks with Renewable Energy Energy Cooperation and Traffic Management in Cellular Networks with Renewable Energy Hyun-Suk Lee Dept. of Electrical and Electronic Eng., Yonsei University, Seoul, Korea Jang-Won Lee Dept. of Electrical

More information

Privacy-Aware Offloading in Mobile-Edge Computing

Privacy-Aware Offloading in Mobile-Edge Computing Privacy-Aware Offloading in Mobile-Edge Computing Xiaofan He Juan Liu Richeng Jin and Huaiyu Dai EE Dept., Lamar University EE Dept., Ningbo University ECE Dept., North Carolina State University e-mail:

More information

2312 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 3, MARCH 2016

2312 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 3, MARCH 2016 2312 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 3, MARCH 2016 Energy-Efficient Resource Allocation for Wireless Powered Communication Networs Qingqing Wu, Student Member, IEEE, Meixia Tao,

More information

Optimal 1D Trajectory Design for UAV-Enabled Multiuser Wireless Power Transfer

Optimal 1D Trajectory Design for UAV-Enabled Multiuser Wireless Power Transfer Optimal 1D Trajectory Design for UA-Enabled Multiuser Wireless Power Transfer Yulin Hu 1, Xiaopeng Yuan 1, Jie Xu 2, and Anke Schmeink 1 1 SEK Research Group, RWTH Aachen University, 5262 Aachen, Germany.

More information

Joint FEC Encoder and Linear Precoder Design for MIMO Systems with Antenna Correlation

Joint FEC Encoder and Linear Precoder Design for MIMO Systems with Antenna Correlation Joint FEC Encoder and Linear Precoder Design for MIMO Systems with Antenna Correlation Chongbin Xu, Peng Wang, Zhonghao Zhang, and Li Ping City University of Hong Kong 1 Outline Background Mutual Information

More information

Cooperative Energy Harvesting Communications with Relaying and Energy Sharing

Cooperative Energy Harvesting Communications with Relaying and Energy Sharing Cooperative Energy Harvesting Communications with Relaying and Energy Sharing Kaya Tutuncuoglu and Aylin Yener Department of Electrical Engineering The Pennsylvania State University, University Park, PA

More information

Network Calculus. A General Framework for Interference Management and Resource Allocation. Martin Schubert

Network Calculus. A General Framework for Interference Management and Resource Allocation. Martin Schubert Network Calculus A General Framework for Interference Management and Resource Allocation Martin Schubert Fraunhofer Institute for Telecommunications HHI, Berlin, Germany Fraunhofer German-Sino Lab for

More information

ABSTRACT WIRELESS COMMUNICATIONS. criterion. Therefore, it is imperative to design advanced transmission schemes to

ABSTRACT WIRELESS COMMUNICATIONS. criterion. Therefore, it is imperative to design advanced transmission schemes to ABSTRACT Title of dissertation: DELAY MINIMIZATION IN ENERGY CONSTRAINED WIRELESS COMMUNICATIONS Jing Yang, Doctor of Philosophy, 2010 Dissertation directed by: Professor Şennur Ulukuş Department of Electrical

More information

Load Regulating Algorithm for Static-Priority Task Scheduling on Multiprocessors

Load Regulating Algorithm for Static-Priority Task Scheduling on Multiprocessors Technical Report No. 2009-7 Load Regulating Algorithm for Static-Priority Task Scheduling on Multiprocessors RISAT MAHMUD PATHAN JAN JONSSON Department of Computer Science and Engineering CHALMERS UNIVERSITY

More information

Enhancing Energy Efficiency among Communication, Computation and Caching with QoI-Guarantee

Enhancing Energy Efficiency among Communication, Computation and Caching with QoI-Guarantee Enhancing Energy Efficiency among Communication, Computation and Caching with QoI-Guarantee Faheem Zafari 1, Jian Li 2, Kin K. Leung 1, Don Towsley 2, Ananthram Swami 3 1 Imperial College London 2 University

More information

Optimal Energy Management Strategies in Wireless Data and Energy Cooperative Communications

Optimal Energy Management Strategies in Wireless Data and Energy Cooperative Communications 1 Optimal Energy Management Strategies in Wireless Data and Energy Cooperative Communications Jun Zhou, Xiaodai Dong, Senior Member, IEEE arxiv:1801.09166v1 [eess.sp] 28 Jan 2018 and Wu-Sheng Lu, Fellow,

More information

Network Flow-based Simultaneous Retiming and Slack Budgeting for Low Power Design

Network Flow-based Simultaneous Retiming and Slack Budgeting for Low Power Design Outline Network Flow-based Simultaneous Retiming and Slack Budgeting for Low Power Design Bei Yu 1 Sheqin Dong 1 Yuchun Ma 1 Tao Lin 1 Yu Wang 1 Song Chen 2 Satoshi GOTO 2 1 Department of Computer Science

More information

Trust Degree Based Beamforming for Multi-Antenna Cooperative Communication Systems

Trust Degree Based Beamforming for Multi-Antenna Cooperative Communication Systems Introduction Main Results Simulation Conclusions Trust Degree Based Beamforming for Multi-Antenna Cooperative Communication Systems Mojtaba Vaezi joint work with H. Inaltekin, W. Shin, H. V. Poor, and

More information

Joint Optimization of Radio Resources and Code Partitioning in Mobile Edge Computing

Joint Optimization of Radio Resources and Code Partitioning in Mobile Edge Computing 1 Joint Optimization of Radio Resources and Code Partitioning in Mobile Edge Computing Paolo Di Lorenzo, Member, IEEE, Sergio Barbarossa, Fellow, IEEE, and Stefania Sardellitti, Member, IEEE, arxiv:137.3835v3

More information

Application of Optimization Methods and Edge AI

Application of Optimization Methods and Edge AI and Hai-Liang Zhao hliangzhao97@gmail.com November 17, 2018 This slide can be downloaded at Link. Outline 1 How to Design Novel Models with Methods Embedded Naturally? Outline 1 How to Design Novel Models

More information

Revenue Maximization in a Cloud Federation

Revenue Maximization in a Cloud Federation Revenue Maximization in a Cloud Federation Makhlouf Hadji and Djamal Zeghlache September 14th, 2015 IRT SystemX/ Telecom SudParis Makhlouf Hadji Outline of the presentation 01 Introduction 02 03 04 05

More information

Achievable Throughput of Energy Harvesting Fading Multiple-Access Channels under Statistical QoS Constraints

Achievable Throughput of Energy Harvesting Fading Multiple-Access Channels under Statistical QoS Constraints Achievable Throughput of Energy Harvesting Fading Multiple-Access Channels under Statistical QoS Constraints Deli Qiao and Jingwen Han Abstract This paper studies the achievable throughput of fading multiple-access

More information

Approximate Queueing Model for Multi-rate Multi-user MIMO systems.

Approximate Queueing Model for Multi-rate Multi-user MIMO systems. An Approximate Queueing Model for Multi-rate Multi-user MIMO systems Boris Bellalta,Vanesa Daza, Miquel Oliver Abstract A queueing model for Multi-rate Multi-user MIMO systems is presented. The model is

More information

ADAPTIVE CLUSTERING ALGORITHM FOR COOPERATIVE SPECTRUM SENSING IN MOBILE ENVIRONMENTS. Jesus Perez and Ignacio Santamaria

ADAPTIVE CLUSTERING ALGORITHM FOR COOPERATIVE SPECTRUM SENSING IN MOBILE ENVIRONMENTS. Jesus Perez and Ignacio Santamaria ADAPTIVE CLUSTERING ALGORITHM FOR COOPERATIVE SPECTRUM SENSING IN MOBILE ENVIRONMENTS Jesus Perez and Ignacio Santamaria Advanced Signal Processing Group, University of Cantabria, Spain, https://gtas.unican.es/

More information

Achievable Throughput of Energy Harvesting Fading Multiple-Access Channels under Statistical QoS Constraints

Achievable Throughput of Energy Harvesting Fading Multiple-Access Channels under Statistical QoS Constraints Achievable Throughput of Energy Harvesting Fading Multiple-Access Channels under Statistical QoS Constraints Deli Qiao and Jingwen Han Abstract This paper studies the achievable throughput of fading multiple-access

More information

Fast Near-Optimal Energy Allocation for Multimedia Loading on Multicarrier Systems

Fast Near-Optimal Energy Allocation for Multimedia Loading on Multicarrier Systems Fast Near-Optimal Energy Allocation for Multimedia Loading on Multicarrier Systems Michael A. Enright and C.-C. Jay Kuo Department of Electrical Engineering and Signal and Image Processing Institute University

More information

Channel Allocation Using Pricing in Satellite Networks

Channel Allocation Using Pricing in Satellite Networks Channel Allocation Using Pricing in Satellite Networks Jun Sun and Eytan Modiano Laboratory for Information and Decision Systems Massachusetts Institute of Technology {junsun, modiano}@mitedu Abstract

More information

Optimal Harvest-or-Transmit Strategy for Energy Harvesting Underlay Cognitive Radio Network

Optimal Harvest-or-Transmit Strategy for Energy Harvesting Underlay Cognitive Radio Network Optimal Harvest-or-Transmit Strategy for Energy Harvesting Underlay Cognitive Radio Network Kalpant Pathak and Adrish Banerjee Department of Electrical Engineering, Indian Institute of Technology Kanpur,

More information

On Two Class-Constrained Versions of the Multiple Knapsack Problem

On Two Class-Constrained Versions of the Multiple Knapsack Problem On Two Class-Constrained Versions of the Multiple Knapsack Problem Hadas Shachnai Tami Tamir Department of Computer Science The Technion, Haifa 32000, Israel Abstract We study two variants of the classic

More information

Decoupling Coupled Constraints Through Utility Design

Decoupling Coupled Constraints Through Utility Design 1 Decoupling Coupled Constraints Through Utility Design Na Li and Jason R. Marden Abstract The central goal in multiagent systems is to design local control laws for the individual agents to ensure that

More information

5682 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 55, NO. 12, DECEMBER /$ IEEE

5682 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 55, NO. 12, DECEMBER /$ IEEE 5682 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 55, NO. 12, DECEMBER 2009 Hyperplane-Based Vector Quantization for Distributed Estimation in Wireless Sensor Networks Jun Fang, Member, IEEE, and Hongbin

More information

Design of Non-Binary Quasi-Cyclic LDPC Codes by Absorbing Set Removal

Design of Non-Binary Quasi-Cyclic LDPC Codes by Absorbing Set Removal Design of Non-Binary Quasi-Cyclic LDPC Codes by Absorbing Set Removal Behzad Amiri Electrical Eng. Department University of California, Los Angeles Los Angeles, USA Email: amiri@ucla.edu Jorge Arturo Flores

More information

Crowdsourcing in Cyber-Physical Systems: Stochastic Optimization with Strong Stability

Crowdsourcing in Cyber-Physical Systems: Stochastic Optimization with Strong Stability Received 1 April 2013; revised 25 May 2013; accepted 29 June 2013. Date of publication 15 July 2013; date of current version 21 January 2014. Digital Object Identifier 10.1109/EC.2013.2273358 Crowdsourcing

More information

Embedded Systems Development

Embedded Systems Development Embedded Systems Development Lecture 3 Real-Time Scheduling Dr. Daniel Kästner AbsInt Angewandte Informatik GmbH kaestner@absint.com Model-based Software Development Generator Lustre programs Esterel programs

More information

Stochastic Control of Computation Offloading to a Helper with a Dynamically Loaded CPU

Stochastic Control of Computation Offloading to a Helper with a Dynamically Loaded CPU Stochastic Control of Computation Offloading to a Helper with a Dynamically Loaded CPU Yunzheng Tao, Changsheng You, Ping Zhang, and aibin Huang arxiv:802.00v [cs.it] 27 Feb 208 Abstract Due to densification

More information

Research on Consistency Problem of Network Multi-agent Car System with State Predictor

Research on Consistency Problem of Network Multi-agent Car System with State Predictor International Core Journal of Engineering Vol. No.9 06 ISSN: 44-895 Research on Consistency Problem of Network Multi-agent Car System with State Predictor Yue Duan a, Linli Zhou b and Yue Wu c Institute

More information

Expectation propagation for signal detection in flat-fading channels

Expectation propagation for signal detection in flat-fading channels Expectation propagation for signal detection in flat-fading channels Yuan Qi MIT Media Lab Cambridge, MA, 02139 USA yuanqi@media.mit.edu Thomas Minka CMU Statistics Department Pittsburgh, PA 15213 USA

More information

Average Throughput Analysis of Downlink Cellular Networks with Multi-Antenna Base Stations

Average Throughput Analysis of Downlink Cellular Networks with Multi-Antenna Base Stations Average Throughput Analysis of Downlink Cellular Networks with Multi-Antenna Base Stations Rui Wang, Jun Zhang, S.H. Song and K. B. Letaief, Fellow, IEEE Dept. of ECE, The Hong Kong University of Science

More information

Fundamental Limits of Cloud and Cache-Aided Interference Management with Multi-Antenna Edge Nodes

Fundamental Limits of Cloud and Cache-Aided Interference Management with Multi-Antenna Edge Nodes Fundamental Limits of Cloud and Cache-Aided Interference Management with Multi-Antenna Edge Nodes Jingjing Zhang and Osvaldo Simeone arxiv:72.04266v4 [cs.it] 2 Mar 208 Abstract In fog-aided cellular systems,

More information

Nash Bargaining in Beamforming Games with Quantized CSI in Two-user Interference Channels

Nash Bargaining in Beamforming Games with Quantized CSI in Two-user Interference Channels Nash Bargaining in Beamforming Games with Quantized CSI in Two-user Interference Channels Jung Hoon Lee and Huaiyu Dai Department of Electrical and Computer Engineering, North Carolina State University,

More information

These outputs can be written in a more convenient form: with y(i) = Hc m (i) n(i) y(i) = (y(i); ; y K (i)) T ; c m (i) = (c m (i); ; c m K(i)) T and n

These outputs can be written in a more convenient form: with y(i) = Hc m (i) n(i) y(i) = (y(i); ; y K (i)) T ; c m (i) = (c m (i); ; c m K(i)) T and n Binary Codes for synchronous DS-CDMA Stefan Bruck, Ulrich Sorger Institute for Network- and Signal Theory Darmstadt University of Technology Merckstr. 25, 6428 Darmstadt, Germany Tel.: 49 65 629, Fax:

More information

6196 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 57, NO. 9, SEPTEMBER 2011

6196 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 57, NO. 9, SEPTEMBER 2011 6196 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 57, NO. 9, SEPTEMBER 2011 On the Structure of Real-Time Encoding and Decoding Functions in a Multiterminal Communication System Ashutosh Nayyar, Student

More information