Research Article An Improved Proportionate Normalized Least-Mean-Square Algorithm for Broadband Multipath Channel Estimation

Size: px
Start display at page:

Download "Research Article An Improved Proportionate Normalized Least-Mean-Square Algorithm for Broadband Multipath Channel Estimation"

Transcription

1 e Scientific World Journal, Article ID 72969, 9 ages htt://dx.doi.org/1.11/214/72969 Research Article An Imroved Proortionate Normalized Least-Mean-Square Algorithm for Broadband Multiath Channel Estimation Yingsong Li and Masanori Hamamura Graduate School of Engineering, Kochi University of Technology, Kami-shi , Jaan Corresondence should be addressed to Yingsong Li; liyingsong@ieee.org Received 27 December 213; Acceted 19 February 214; Published 2 March 214 Academic Editors: H. R. Karimi, X. Yang, Z. Yu, and W. Zhang Coyright 214 Y. Li and M. Hamamura. This is an oen access article distributed under the Creative Commons Attribution License, which ermits unrestricted use, distribution, and reroduction in any medium, rovided the original work is roerly cited. To make use of the sarsity roerty of broadband multiath wireless communication channels, we mathematically roose an l - norm-constrained roortionate normalized least-mean-square (-) sarse channel estimation algorithm. A general l - norm is weighted by the gain matrix and is incororated into the cost function of the roortionate normalized least-mean-square () algorithm. This integration is equivalent to adding a zero attractor to the iterations, by which the convergence seed and steady-state erformance of the inactive tas are significantly imroved. Our simulation results demonstrate that the roosed algorithm can effectively imrove the estimation erformance of the -based algorithm for sarse channel estimation alications. 1. Introduction Broadband signal transmission is becoming a commonly used high-data-rate technique for next-generation wireless communication systems, such as 3 GPP long-term evolution (LTE) and worldwide interoerability for microwave access (WiMAX) [1]. The transmission erformance of coherent detection for such broadband communication systems strongly deends on the quality of channel estimation [2 ]. Fortunately, broadband multiath channels can be accurately estimated using adative filter techniques [6 1] suchas the normalized least-mean-square () algorithm, which has low comlexity and can be easily imlemented at the receiver. On the other hand, channel measurements have shown that broadband wireless multiath channels can often be described by only a small number of roagation aths with long delays [4, 11, 12]. Thus, a broadband multiath channel can be regarded as a sarse channel with only a few active dominant tas, while the other inactive tas are zero or close to zero. This inherent sarsity of the channel imulse resonse (CIR) can be exloited to imrove the quality of channel estimation. However, such classical algorithms with a uniform ste size across all filter coefficients have slow convergence when estimating sarse imulseresonsesignalssuchasthoseinbroadbandsarse wireless multiath channels [11]. Consequently, corresonding algorithms have recently received significant attention in the context of comressed sensing (CS) [, 12 14] and were already considered for channel estimation rior to the CS era [, 12]. However, these CS channel estimation algorithms are sensitivetothenoiseinwirelessmultiathchannels. Insired by the CS theory [12 14], several zero-attracting (ZA) algorithms have been roosed and investigated by combining the CS theory and the standard least-meansquare (LMS) algorithm for echo cancellation and system identification, which are known as the zero-attracting LMS (ZA-LMS) and reweighted ZA-LMS (RZA-LMS) algorithms, resectively [1]. Recently, this technique has been exanded to the algorithm and other adative filter algorithms to imrove their convergence seed in a sarse environment [9, 16 18]. However, these aroaches are mainly designed for nonroortionate adative algorithms. On the other hand, to utilize the advantages of the algorithm, such as stable erformance and low comlexity, the roortionate

2 2 The Scientific World Journal normalized least-mean-square () algorithm has been roosedandstudiedtoexloitthesarsityinnature[19]and has been alied to echo cancellation in telehone networks. Although the algorithm can utilize the sarsity characteristics of a sarse signal and obtain faster convergence at the initial stage by assigning indeendent magnitudes to the active tas, the convergence seed is reduced by even more than that of the algorithm for the inactive tas after theactivetasconverge.consequently,severalalgorithms have been roosed to imrove the convergence seed of the algorithm [2 27], which include the use of the l 1 -norm technique and a variable ste size. Although thesealgorithmshavesignificantlyimrovedtheconvergence seed of the algorithm, they still converge slowly after the active tas converge. In addition, some of them are inferior to the and algorithms in terms of the steady-state error when the sarsity decreases. From these reviously roosed sarse signal estimation algorithms, we know that the ZA algorithms mainly exert a enalty on the inactive channel tas through the integration of the l 1 - norm constraint into the cost function of the standard LMS algorithms to achieve better estimation erformance, while the algorithm udates each filter coefficient with an indeendent ste size, which imroves the convergence of the active tas. Motivated by the CS theory [13, 14] andzatechnique [1 18], we roose an l -norm-constrained (- ) algorithm that incororates the l -norm into the cost function of the algorithm, resulting in an imroved roortionate adative algorithm. The difference between the roosed - algorithm and the ZA algorithms is that the gain-matrix-weighted l -norm is used in our roosed - algorithm instead of the general l 1 -norm to exand the alication of ZA algorithms [1]. Also, this integration is equivalent to adding a zero attractor in the iterations of the algorithm to obtain the benefits of both the and ZA algorithms. Thus, our roosed - algorithm can achieve fast convergence attheinitialstagefortheactivetas.aftertheconvergence of these active tas, the ZA technique in the - algorithm acts as another force to attract the inactive tas to zero to arrest the slow convergence of the algorithm. Furthermore, our roosed - algorithm achieves a lower mean square error than the algorithm and its related imroved algorithms, such as the imroved (I) [2] andμ-law (M) [21] algorithms. In this study, our roosed - algorithm is verified over a sarse multiath channel by comarison with the,, I, and M algorithms. The simulation results demonstrate that the - algorithm achieves better channel estimation erformance in terms of both convergence seed and steady-state behavior for sarse channel estimation. The remainder of this aer is organized as follows. Section 2 briefly reviews the standard,, andimrovedalgorithms,includingthei and M algorithms. In Section 3, we describe in detail the roosed - algorithm, which emloys the Lagrange multilier method. In Section4, the estimation erformance of the roosed - algorithm is verified over sarse channels and comared with other commonly used algorithms. Finally, this aer is concluded in Section. 2. Related Channel Estimation Algorithms 2.1. Normalized Least-Mean-Square Algorithm. In this section, we first consider the sarse multiath communication system shown in Figure 1 to discuss the channel estimation algorithms. The inut signal x(n)=[x(n),x(n 1),...,x(n N+1)] T containing the N most recent samles is transmitted over a finite imulse resonse (FIR) channel with channel imulse resonse (CIR) h = [h,h 1,...,h N 1 ] T,where( ) T denotes the transosition oeration. Then the outut signal of the channel is written as follows: y (n) = h T x (n), (1) where h is a sarse channel vector with K dominant active taswhosemagnitudesarelargerthanzeroand(n K) inactive tas whose magnitudes are zero or close to zero with K N.Toestimatetheunknownsarsechannelh,an algorithm uses the inut signal x(n), the outut signal y(n), and the instantaneous estimation error e(n), which is given by e (n) =d(n) ĥt (n) x (n), (2) where ĥ(n) istheadativechannelestimatoratinstant n, d(n) = y(n) + V(n),andV(n) is an additive noise at the receiver. The udate function of the channel estimation algorithm is exressed as ĥ (n+1) = ĥ (n) +μ e (n) x (n) x T, (3) (n) x (n) +δ where μ is the ste size with <μ <2and δ is a small ositive constant used to avoid division by zero Proortionate Normalized Least-Mean-Square Algorithm. The algorithm, which is an algorithm imroved by the use of a roortionate technique, has been roosed for sarse system identification and echo cancellation. In this algorithm, each ta is assigned an individual ste size, which is obtained from the revious estimation of the filter coefficient. According to the gain allocation rule in this algorithm, the greater the magnitude of the ta, the larger the ste size assigned to it, and hence the active tas converge quickly. The udate function of the algorithm [19] is described by the following equation with reference to Figure 1: ĥ (n+1) = ĥ (n) +μ e (n) G (n) x (n) x T. (n) G (n) x (n) +δ (4) Here, G(n), which denotes as the gain matrix, is a diagonal matrix that modifies the ste size of each ta, μ is the

3 The Scientific World Journal 3 Inut signal x(n) Unknown FIR channel h Estimated FIR channel ĥ(n) Adative algorithms Adative channel estimation Additive noise Outut signal + (n) y(n) + ŷ(n) e(n) + d(n) Figure 1: Tyical sarse multiath communication system. global ste size of the algorithm, and δ =δ 2 x /N is a regularization arameter to revent division by zero at the initialization stage, where δ 2 x is the ower of the inut signal x(n). In the algorithm, the gain matrix G(n) is given by G (n) = diag (g (n),g 1 (n),...,g N 1 (n)), () where the individual gain g i (n) is defined as with g i (n) = γ i (n) N 1 i= γ i (n), i N 1 (6) γ i (n) = max [ρ g max [δ, ĥ (n), ĥ1 (n),..., ĥn 1 (n) ], ĥi (n) ], (7) where the arameters δ and ρ g are ositive constants with tyical values of δ =.1 and ρ g = /N. δ is used to regularize the udating at the initial stage when all the tas areinitializedtozero,andρ g is used to revent ĥi(n) from stalling when it is much smaller than the largest coefficient Imroved Proortionate Normalized Least-Mean-Square Algorithms I Algorithm. The I algorithm is a tye of algorithm used to imrove the convergence seed of the algorithm. It is a combination of the and algorithms with the relative significance of each coefficient controlled by a factor α. The I algorithm [2] adotsthel 1 -norm to enable the smooth selection of (7), and the udate equation of the I algorithm is exressed as ĥ (n+1) = ĥ (n) μ e (n) K (n) x (n) I x T, (n) K (n) x (n) +δ I (8) where K(n) = diag(k (n), k 1 (n),...,k N 1 (n)) is a diagonal matrix used to adjust the ste size of the I algorithm, where k j (n) = 1 α 2N + (1+α) (n) ĥj 2 ĥ (n) +ε, 1 j N 1 for a small ositive constant ε and 1 α 1.Attheinitial stage,thestesizeismultiliedby(1 α)/2n, sinceallthe filter coefficients are initialized to zero. Thus, in the I algorithm, a regularization arameter δ I is introduced, which is given by (9) δ I = 1 α 2N δ. (1) We can see that the I is identical to the algorithmforα = 1, while the I behaves identically to the algorithm when α=1. In ractical engineering alications, a suitable value for α is or M Algorithm. The μ-law algorithm (M) is another enhancement of the algorithm that utilizes the logarithm of the magnitudes of the filter coefficients instead of using the magnitudes directly in the algorithm [21]. The udate equation is the same as that in the algorithm given by (4). In the M algorithm, γ i (n) = max [ρ g max [δ F( ĥ (n) ),F( ĥ1 (n) ),..., where F( ĥn 1 (n) )], F( ĥi (n) )], (11) F( (n) ĥi )=log (1 + θ (n) ĥi ), (12) where θ is a large ositive constant related to the estimation accuracy requirement, tyically θ = Proosed - Algorithm In this section, we roose an - algorithm by incororating the l -norm into the cost function of the algorithm to create a zero attractor, making it a tye of ZA algorithm. The difference between the - algorithm and general ZA algorithms is that the gain-matrixweighted l -norm is taken into account in designing the zero attractor. On the other hand, the roosed - algorithm is based on the commonly used algorithm, which is also a sarse channel estimation algorithm and can imrove the convergence for the active tas. Regarding

4 4 The Scientific World Journal channel estimation, the urose of the - algorithm is to minimize (ĥ (n+1) ĥ (n))t G 1 (n) where ε is a small value to revent division by zero. By multilying both sides of (17)byx T (n),weobtain x T (n) ĥ (n+1) = xt (n) ĥ (n) +λxt (n) G (n) x (n) (ĥ (n+1) ĥ (n))+γ G 1 (n)ĥ(n + 1) subject to d (n) ĥt (n+1) x (n) =, (13) 1 x T (n) ĥ(n) γ. 1 ĥ(n) +ε (18) where G 1 (n) is the inverse of the gain matrix G(n) in the algorithm, γ >is a very small constant used to balance the estimation error and the sarse l -norm enalty of ĥ(n + 1), is the -norm defined as ĥ =( i ĥ i )1/, and 1.Notethatin(13), we introduce an l -norm enalty to ĥ(n + 1) after scaling the gain matrix by G 1 (n), which is different from the reviously roosed ZA LMS algorithms. To minimize (13),theLagrangemultiliermethodis adoted, and the cost function J (n + 1) of the roosed - algorithm is exressed as J (n+1) = (ĥ (n+1) ĥ (n))t G 1 (n) (ĥ (n+1) ĥ (n))+γ G 1 ĥ(n + 1) +λ (d (n) ĥt (n+1) x (n)), (14) where λ isthelagrangemultilier. By calculating the gradient of the cost function J (n + 1) of the - algorithm and assuming ĥ(n + 1) = ĥ(n) in the steady stage, we have J (n+1) ĥ (n+1) =, J (n+1) λ =, (1) ĥ (n+1) = ĥ (n) +λg (n) x (n) γ ĥ(n). 1 ĥ(n) (16) 1 In ractice, we need to introduce a small ositive constant into the final term in (16) tocoewiththesituationthatan entry of ĥ(n) aroaches zero, which is the case for a sarse CIR at initialization. Then the udate equation (16)ofthe- algorithm is modified to ĥ (n+1) = ĥ (n) +λg (n) x (n) γ ĥ(n), 1 ĥ(n) +ε (17) 1 From (2), (1), and (17), we obtain e (n) = γ x T (n) ĥ(n) 1 ĥ(n) 1 +ε +λx T (n) G (n) x (n). (19) Then, the Lagrange multilier λ is given as follows by solving (19): 1 x T (n) ĥ(n) λ= (e(n) +γ ) 1 ĥ(n) +ε (x T (n) G (n) x (n)) 1. Substituting (2)into(17), we have 1 ĥ (n+1) = ĥ (n) γ ĥ(n) 1 ĥ(n) +ε = 1 (2) x T (n) ĥ(n) + ((e (n) +γ ) 1 ĥ(n) +ε (x T (n)g(n)x(n)) 1 ) G (n) x (n) ĥ (n) + e (n) G (n) x (n) x T (n) G (n) x (n) γ {I G (n) x (n) xt (n) x T (n) G (n) x (n) } 1 ĥ (n). ĥ (n) 1 +ε (21) It was found that the magnitudes of the elements in the matrix G(n)x(n)x T (n){x T (n)g(n)x(n)} 1 are much smaller

5 The Scientific World Journal than 1 for broadband multiath channel estimation. Therefore, the udate equation (21) of the roosed - algorithm is rewritten as ĥ (n+1) = ĥ (n) + e (n) G (n) x (n) x T (n) G (n) x (n) 1 γ ĥ(n). 1 ĥ(n) +ε (22) Here, we neglect the effects of the matrix G(n)x(n)x T (n) {x T (n)g(n)x(n)} 1 and assume that the filter order is large. Similarly to the algorithm, a ste size μ is introduced to balance the convergence seed and the steady-state error of the roosed - algorithm, and a small ositive constant ε =δ 2 x /N is emloyed to revent division byzero.thus,theudatefunction(22)canbemodifiedto ĥ (n+1) = ĥ (n) +μ e (n) G (n) x (n) x T (n) G (n) x (n) +ε ρ ĥ(n) 1 ĥ(n) 1 +ε = ĥ (n) +μ e (n) G (n) x (n) x T ρ (n) G (n) x (n) +ε T (n), (23) where ρ =μ γ and T(n)= ĥ(n) 1 sgn(ĥ(n)){ ĥ(n) 1 + ε } 1.Comaringtheudatefunction(23) oftheroosed - algorithm with the udate function (4) ofthe algorithm, we see that our roosed - algorithm has the additional term γ T(n), also defined as the zero attractor, which attracts the small channel tas to zero with high robability. Moreover, the ZA strength of this zero attractor is controlled by ρ. In other words, in our roosed - algorithm, the gain matrix G(n) assigns a large stesizetotheactivechanneltasofthesarsechannel,while the zero attractor mainly exerts the l -enalty on the inactive tas whose tas are zero or close to zero. Thus, our roosed - algorithm can further imrove the convergence seed of the algorithm after the convergence of the large active tas. 4. Results and Discussions In this section, we resent the results of comuter simulations carried out to illustrate the channel estimation erformance oftheroosed-algorithmoverasarsemultiath communication channel and comare it with those of the reviously roosed I, M,, and algorithms. Here, we consider a sarse channel h whose length N is 64 or 128 and whose number of dominant active tas K is set to three different sarsity levels, namely, K=2,4and 8, similar to revious studies [6, 22, 2, 26]. The dominant active channel tas are obtained from a Gaussian Magnitude Channel ta index Figure 2: Tyical sarse multiath channel. distribution with h 2 2 =1, and the ositions of the dominant channel tas are randomly saced along the length of the channel. The inut signal x(n) of the channel is a Gaussian random signal while the outut of the channel is corruted by an indeendent white Gaussian noise V(n). An examle ofatyicalsarsemultiathchannelwithachannellength of N = 64 and a sarsity level of K = 3 is shown in Figure 2.Inthesimulations,theowerofthereceivedsignal is E b =1,whilethenoiseowerisgivenbyδ 2 V and the signal -to-noise ratio is defined as SNR = 1 log(e b /δ 2 V ).Inallthe simulations, the difference between the actual and estimated channels based on the sarsity-aware algorithms and the sarse channel mentioned above is evaluated by the MSE defined as follows: MSE (n) =1log 1 E { h ĥ(n) } (db). (24) 2 In these simulations, the simulation arameters are chosen to be μ =μ =μ I =μ =., δ =.1, ε =.1, α=, ε =., ρ =1 1, δ =.1, ρ g =/N, θ = 1, =., andsnr=3db.whenwe change one of these arameters, the other arameters remain constant Estimation Performance of the Proosed - Algorithm Effects of Parameters on the Proosed - Algorithm. Intheroosed-algorithm,therearetwo extra arameters, and ρ,comaredwiththe algorithm, which are introduced to design the zero attractor. Next, we show how these two arameters affect the roosed - algorithm over a sarse channel with N=64or 128 and K = 4. The simulation results for different values of ρ and are shown in Figures 3 and 4, resectively. From Figure 3(a),wecanseethatthesteady-stateerrorofthe - algorithm decreases with decreasing ρ when 2

6 6 The Scientific World Journal ρ =4 1 ρ =3 1 ρ =2 1 ρ =1 1 ρ =2 1 6 ρ =2 1 7 ρ =4 1 ρ =3 1 ρ =2 1 ρ =1 1 ρ =2 1 6 ρ =2 1 7 (a) N=64 (b) N = 128 Figure 3: Effects of ρ on the roosed - algorithm. ρ 2 1 6, while it increases again when ρ is less than Furthermore, the convergence seed of the - algorithm raidly decreases when ρ is less than 1 1. This is because a small ρ results in a low ZA strength, which consequently reduces the convergence seed. In the case of N = 128 shown in Figure 3(b),weobservethat both the convergence seed and the steady-state erformance are imroved with decreasing ρ for ρ 1 1.When ρ < 1 1,theconvergenceseedofthe- algorithm decreases while the steady-state error remains constant. Figure4 demonstrates the effects of the arameter. We can see from Figure 4(a) that the convergence seed of the roosed - algorithm raidly decreases with increasing for N = 64. Moreover, the steady-state error is reduced with ranging from.4 to., while it remains constant for =.6,.7, and.8. However, the steadystate erformance for =1isinferior to that for =.8. Thisisbecausetheroosed-algorithmisan l 1 -norm-enalized algorithm, which cannot distinguish between active tas and inactive tas, reducing its convergence seed and steady-state erformance. When N= 128, asshowninfigure 4(b), the steady-state erformance is imroved as increases from.4 to.6. Thus, we should carefully select the arameters ρ and to balance the convergence seed and steady-state erformance for the roosed - algorithm EffectsofSarsityLevelontheProosed-Algorithm. On the basis of the results discussed in Section for our roosed - algorithm, we choose =. and ρ = 1 1 to evaluate the channel estimation erformance of the - algorithm over a sarse channel with different channel lengths of N=64and 128, for which the obtained simulation results are given in Figures and 6, resectively.fromfigure, weseethatour roosed - algorithm has the same convergence seed as the algorithm at the initial stage. The roosed - algorithm converges faster than the algorithm as well as the I and algorithms for all sarsity levels K, while its convergence is slightly slower than that of the M algorithm before it reaches a steady stage. However, the roosed - algorithm has the smallest steady-state error for N = 64. When N = 128, we see from Figure 6 that our roosed - algorithm not only has the highest convergence seedbutalsoossessesthebeststeady-stateerformance. This is because with increasing sarsity, our roosed - algorithm attracts the inactive tas to zero quickly and hence the convergence seed is significantly imroved, while the reviously roosed algorithms mainly adjust the ste size of the active tas and thus they only imact on the convergence seed at the early iteration stage. Additionally, we see from Figures and 6 that both theconvergenceseedandthesteady-stateerformanceof all the algorithms deteriorate when the sarsity level K increases for both N = 64 and 128. Inarticular, when K = 8, the convergence seeds of the and I algorithms are greater than that of the algorithm at the early iteration stage, while after this fast initial convergence, their convergence seeds decrease to less than that of the algorithm before reaching a steady

7 The Scientific World Journal =.4 =. =.6 =.7 =.8 =1 =.4 =. =.6 =.7 =.8 =1 (a) N=64 (b) N = 128 Figure 4: Effects of on the roosed - algorithm. stage. Furthermore, we observe that the M algorithm is sensitive to the length N of the channel, and its convergence seed for N = 128 is less than that for N=64at the same sarsity level K andlessthanthatoftheroosed- algorithm. Thus, we conclude that our roosed - algorithm is suerior to the reviously roosed algorithms in terms of both the convergence seed and the steady-state erformance with the aroriate selection of the related arameters and ρ.fromtheabovediscussion, we believe that the gain-matrix-weighted l -norm method inthe-algorithmcanbeusedtofurtherimrove the channel estimation erformance of the I and M algorithms Comutational Comlexity. Finally, we discuss the comutational comlexity of the roosed - algorithm and comare it with those of the,, I, and M algorithms. Here, the comutational comlexity is the arithmetic comlexity, which includes additions, multilications, and divisions. The comutational comlexities of the roosed - algorithm and the related and algorithms are shown in Table 1. From Table 1, we see that the comutational comlexity of our roosed - algorithm is slightly higher than those of the M and algorithms, which is due to the calculation of the gradient of the l -norm. Furthermore, the M algorithm has an additional logarithm oeration, which increases its comlexity but is not included in Table 1. However, the - algorithm noticeably increases the convergence seed and significantly Table 1: Comutational comlexity. Algorithms Additions Multilications Divisions 3N 3N N+3 6N+3 N+2 I 4N + 7 N + N + 2 M N+3 7N+3 N+3-4N+4 9N+4 2N+2 imroves the steady-state erformance of the algorithm. In addition, it also has a higher convergence seed and lower steady-state error than the I and M algorithms when the channel length is large.. Conclusion In this aer, we have roosed an - algorithm to exloit the sarsity of broadband multiath channels and to imroveboththeconvergenceseedandsteady-stateerformance of the algorithm. This algorithm was mainly develoed by incororating the gain-matrix-weighted l - norm into the cost function of the algorithm, which significantly imroves its convergence seed and steady-state erformance. The simulation results demonstrated that our roosed - algorithm, which has an accetable increase in comutational comlexity, increases the convergenceseedandreducesthesteady-stateerrorcomaredwith the reviously roosed algorithms.

8 8 The Scientific World Journal I M - I M - I M - I M - (a) K=2 (b) K=4 (a) K=2 (b) K= I M - (c) K=8 Figure : Effects of sarsity on the roosed - algorithm for N= I M - (c) K=8 Figure 6: Effects of sarsity on the roosed - algorithm for N = 128.

9 The Scientific World Journal 9 Conflict of Interests The authors declare that there is no conflict of interests regarding the ublication of this aer. References [1] L. Korowajczuk, LTE, WiMAX and WLAN Network Design, Otimization and Performance Analysis, JohnWiley&Sons, New York, NY, USA, 211. [2] J. G. Proakis, Digital Communications, McGraw-Hill,4thedition, 21. [3] F.Adachi,D.Grag,S.Takaoka,andK.Takeda, Newdirectionof broadband CDMA techniques, Wireless Communications and Mobile Comuting,vol.7,no.8, ,27. [4] S.F.CotterandB.D.Rao, Sarsechannelestimationviamatching ursuit with alication to equalization, IEEE Transactions on Communications,vol.,no.3, ,22. []W.U.Bajwa,J.Haut,A.M.Sayeed,andR.Nowak, Comressed channel sensing: a new aroach to estimating sarse multiath channels, Proceedings of the IEEE,vol.98,no.6, , 21. [6] H.Zayyani,M.Babaie-Zadeh,andC.Jutten, Comressedsensing block MAP-LMS adative filter for sarse channel estimation and a Bayesian Cramer-Rao bound, in IEEE International Worksho on Machine Learning for Signal Processing (MLSP 9),. 1 6, Grenoble, France, Setember 29. [7] Md. M. Rana, J. Kim, and W.-K. Cho, LMS based channel estimation of LTE ulink using variable ste size and hase information, Radioengineering, vol. 19, no. 4, , 21. [8] J. Homer, I. Mareels, and C. Hoang, Enhanced detection-guided estimation of sarse FIR-modeled signal channels, IEEE Transactions on Circuits and Systems I, vol.3,no.8, , 26. [9] B. Babadi, N. Kaloutsidis, and V. Tarokh, SPARLS: the sarse RLS algorithm, IEEE Transactions on Signal Processing, vol. 8, no. 8, , 21. [1] A. H. Sayed, Adative Filters,JohnWiley&Sons, 28. [11] G. Gui and F. Adachi, Imroved least mean square algorithm with alication to adative sarse channel estimation, EUR- ASIP Journal on Wireless Communications and Networking,vol. 213, article 24, 213. [12]P.Maechler,P.Greisen,B.Sorrer,S.Steiner,N.Felber,and A. Burg, Imlementation of greedy algorithms for LTE sarse channel estimation, in Conference Record of the Forty Fourth Asilomar Conference on Signals, Systems and Comuters (ASI- LOMAR 1),. 4 4, Pacific Grove, Calif, USA, November 21. [13] R. Tibshirani, Regression shrinkage and selection via the lasso, JournaloftheRoyalStatisticalSocietyB,vol.8,no.1, , [14] D. L. Donoho, Comressed sensing, IEEE Transactions on Information Theory,vol.2,no.4, ,26. [1] Y. Chen, Y. Gu, and A. O. Hero III, Sarse LMS for system identification, in Proceedings of the IEEE International Conference on Acoustics, Seech, and Signal Processing (ICASSP 9), , Taiei, Taiwan, Aril 29. [16] O. Taheri and S. A. Vorobyov, Sarse channel estimation with L P -norm and reweighted L 1 -norm enalized least mean squares, in Proceedingsofthe36thIEEEInternationalConference on Acoustics, Seech, and Signal Processing (ICASSP 11), , Prague, Czech Reublic, May 211. [17] Y.Gu,J.Jin,and S.Mei, l norm constraint LMS algorithm for sarse system identification, IEEE Signal Processing Letters,vol. 16, no. 9, , 29. [18] R. L. Das and M. Chakraborty, A zero attracting roortionate normalized least mean square algorithm, in Proceedings of the 2nd Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC 12),.1 4, Hollywood, Calif, USA, December 212. [19] D. L. Duttweiler, Proortionate normalized least-meansquares adatation in echo cancelers, IEEE Transactions on Seech and Audio Processing,vol.8,no.,.8 18,2. [2] J. Benesty and S. L. Gay, An imroved algorithm, in Proceedings of the IEEE International Conference on Acoustic, Seech and Signal Processing, , Orlando, Fla, USA, May 22. [21] H. Deng and M. Doroslovački, Imroving convergence of the algorithm for sarse imulse resonse identification, IEEE Signal Processing Letters,vol.12,no.3, ,2. [22] A. W. H. Khong and P. A. Naylor, Efficient use of sarse adative filters, in Proceedings of the 4th Asilomar Conference on Signals, Systems, and Comuters (ACSSC 6), ,Pacific Grove, Calif, USA, November 26. [23] L. Liu, M. Fukumoto, S. Saiki, and S. Zhang, A variable stesize roortionate algorithm for identification of sarse imulse resonse, IEICE Transactions on Fundamentals of Electronics, Communications and Comuter Sciences, vol.93,no.1, , 21. [24] M. Nekuii and M. Atarodi, A fast converging algorithm for network echo cancelation, IEEE Signal Processing Letters, vol. 11, no. 4, , 24. [2] L. Liu, M. Fukumoto, and S. Saikfi, An imroved mu-law roortionate algorithm, in Proceedings of the IEEE International Conference on Acoustics, Seech and Signal Processing (ICASSP 8), , Las Vegas, Nev, USA, Aril 28. [26] C. Paleologu, J. Benesty, and S. Ciochinǎ, Sarseadative filters for echo cancellation, Synthesis Lectures on Seech and Audio Processing,vol.6,.1 124,21. [27] K. Dogancay, Partial-Udate Adative Signal Processing: Design, Analysis and Imlementation, Academic Press, Oxford, UK, 28.

10 Rotating Machinery Engineering Journal of The Scientific World Journal Distributed Sensor Networks Journal of Sensors Journal of Control Science and Engineering Advances in Civil Engineering Submit your manuscrits at Journal of Journal of Electrical and Comuter Engineering Robotics VLSI Design Advances in OtoElectronics Navigation and Observation Chemical Engineering Active and Passive Electronic Comonents Antennas and Proagation Aerosace Engineering Modelling & Simulation in Engineering Shock and Vibration Advances in Acoustics and Vibration

Proportionate Minimum Error Entropy Algorithm for Sparse System Identification

Proportionate Minimum Error Entropy Algorithm for Sparse System Identification Entroy 05, 7, 5995-6006; doi:0.3390/e7095995 Letter OPEN ACCESS entroy ISSN 099-4300 www.mdi.com/journal/entroy Proortionate Minimum Error Entroy Algorithm for Sarse System Identification Zongze Wu, Siyuan

More information

A Recursive Block Incomplete Factorization. Preconditioner for Adaptive Filtering Problem

A Recursive Block Incomplete Factorization. Preconditioner for Adaptive Filtering Problem Alied Mathematical Sciences, Vol. 7, 03, no. 63, 3-3 HIKARI Ltd, www.m-hiari.com A Recursive Bloc Incomlete Factorization Preconditioner for Adative Filtering Problem Shazia Javed School of Mathematical

More information

Sparsity Promoting LMS for Adaptive Feedback Cancellation

Sparsity Promoting LMS for Adaptive Feedback Cancellation 7 5th Euroean Signal Processing Conference (EUSIPCO) Sarsity Promoting MS for Adative Feedback Cancellation Ching-Hua ee, Bhaskar D. Rao, and Harinath Garudadri Deartment of Electrical and Comuter Engineering

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

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

Online homotopy algorithm for a generalization of the LASSO

Online homotopy algorithm for a generalization of the LASSO This article has been acceted for ublication in a future issue of this journal, but has not been fully edited. Content may change rior to final ublication. Online homotoy algorithm for a generalization

More information

Research Letter An Algorithm to Generate Representations of System Identification Errors

Research Letter An Algorithm to Generate Representations of System Identification Errors Research Letters in Signal Processing Volume 008, Article ID 5991, 4 pages doi:10.1155/008/5991 Research Letter An Algorithm to Generate Representations of System Identification Errors Wancheng Zhang and

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

Adaptive sparse algorithms for estimating sparse channels in broadband wireless communications systems

Adaptive sparse algorithms for estimating sparse channels in broadband wireless communications systems Wireless Signal Processing & Networking Workshop: Emerging Wireless Technologies, Sendai, Japan, 28 Oct. 2013. Adaptive sparse algorithms for estimating sparse channels in broadband wireless communications

More information

Recent Developments in Multilayer Perceptron Neural Networks

Recent Developments in Multilayer Perceptron Neural Networks Recent Develoments in Multilayer Percetron eural etworks Walter H. Delashmit Lockheed Martin Missiles and Fire Control Dallas, Texas 75265 walter.delashmit@lmco.com walter.delashmit@verizon.net Michael

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

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

Iterative Methods for Designing Orthogonal and Biorthogonal Two-channel FIR Filter Banks with Regularities

Iterative Methods for Designing Orthogonal and Biorthogonal Two-channel FIR Filter Banks with Regularities R. Bregović and T. Saramäi, Iterative methods for designing orthogonal and biorthogonal two-channel FIR filter bans with regularities, Proc. Of Int. Worsho on Sectral Transforms and Logic Design for Future

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

Sparse Least Mean Square Algorithm for Estimation of Truncated Volterra Kernels

Sparse Least Mean Square Algorithm for Estimation of Truncated Volterra Kernels Sparse Least Mean Square Algorithm for Estimation of Truncated Volterra Kernels Bijit Kumar Das 1, Mrityunjoy Chakraborty 2 Department of Electronics and Electrical Communication Engineering Indian Institute

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

DSP IC, Solutions. The pseudo-power entering into the adaptor is: 2 b 2 2 ) (a 2. Simple, but long and tedious simplification, yields p = 0.

DSP IC, Solutions. The pseudo-power entering into the adaptor is: 2 b 2 2 ) (a 2. Simple, but long and tedious simplification, yields p = 0. 5 FINITE WORD LENGTH EFFECTS 5.4 For a two-ort adator we have: b a + α(a a ) b a + α(a a ) α R R R + R The seudo-ower entering into the adator is: R (a b ) + R (a b ) Simle, but long and tedious simlification,

More information

Research Article A Note on the Modified q-bernoulli Numbers and Polynomials with Weight α

Research Article A Note on the Modified q-bernoulli Numbers and Polynomials with Weight α Abstract and Alied Analysis Volume 20, Article ID 54534, 8 ages doi:0.55/20/54534 Research Article A Note on the Modified -Bernoulli Numbers and Polynomials with Weight α T. Kim, D. V. Dolgy, 2 S. H. Lee,

More information

Research Article New Mixed Exponential Sums and Their Application

Research Article New Mixed Exponential Sums and Their Application Hindawi Publishing Cororation Alied Mathematics, Article ID 51053, ages htt://dx.doi.org/10.1155/01/51053 Research Article New Mixed Exonential Sums and Their Alication Yu Zhan 1 and Xiaoxue Li 1 DeartmentofScience,HetaoCollege,Bayannur015000,China

More information

Minimax Design of Nonnegative Finite Impulse Response Filters

Minimax Design of Nonnegative Finite Impulse Response Filters Minimax Design of Nonnegative Finite Imulse Resonse Filters Xiaoing Lai, Anke Xue Institute of Information and Control Hangzhou Dianzi University Hangzhou, 3118 China e-mail: laix@hdu.edu.cn; akxue@hdu.edu.cn

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

Multivariable Generalized Predictive Scheme for Gas Turbine Control in Combined Cycle Power Plant

Multivariable Generalized Predictive Scheme for Gas Turbine Control in Combined Cycle Power Plant Multivariable Generalized Predictive Scheme for Gas urbine Control in Combined Cycle Power Plant L.X.Niu and X.J.Liu Deartment of Automation North China Electric Power University Beiing, China, 006 e-mail

More information

ON OPTIMIZATION OF THE MEASUREMENT MATRIX FOR COMPRESSIVE SENSING

ON OPTIMIZATION OF THE MEASUREMENT MATRIX FOR COMPRESSIVE SENSING 8th Euroean Signal Processing Conference (EUSIPCO-2) Aalborg, Denmark, August 23-27, 2 ON OPTIMIZATION OF THE MEASUREMENT MATRIX FOR COMPRESSIVE SENSING Vahid Abolghasemi, Saideh Ferdowsi, Bahador Makkiabadi,2,

More information

AKRON: An Algorithm for Approximating Sparse Kernel Reconstruction

AKRON: An Algorithm for Approximating Sparse Kernel Reconstruction : An Algorithm for Aroximating Sarse Kernel Reconstruction Gregory Ditzler Det. of Electrical and Comuter Engineering The University of Arizona Tucson, AZ 8572 USA ditzler@email.arizona.edu Nidhal Carla

More information

Covariance Matrix Estimation for Reinforcement Learning

Covariance Matrix Estimation for Reinforcement Learning Covariance Matrix Estimation for Reinforcement Learning Tomer Lancewicki Deartment of Electrical Engineering and Comuter Science University of Tennessee Knoxville, TN 37996 tlancewi@utk.edu Itamar Arel

More information

Provided by the author(s) and University College Dublin Library in accordance with ublisher olicies. Please cite the ublished version when available. Title Low Comlexity Stochastic Otimization-Based Model

More information

Discrete Signal Reconstruction by Sum of Absolute Values

Discrete Signal Reconstruction by Sum of Absolute Values JOURNAL OF L A TEX CLASS FILES, VOL., NO. 4, DECEMBER 22 Discrete Signal Reconstruction by Sum of Absolute Values Masaaki Nagahara, Senior Member, IEEE, arxiv:53.5299v [cs.it] 8 Mar 25 Abstract In this

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

Robust Beamforming via Matrix Completion

Robust Beamforming via Matrix Completion Robust Beamforming via Matrix Comletion Shunqiao Sun and Athina P. Petroulu Deartment of Electrical and Comuter Engineering Rutgers, the State University of Ne Jersey Piscataay, Ne Jersey 8854-858 Email:

More information

Tests for Two Proportions in a Stratified Design (Cochran/Mantel-Haenszel Test)

Tests for Two Proportions in a Stratified Design (Cochran/Mantel-Haenszel Test) Chater 225 Tests for Two Proortions in a Stratified Design (Cochran/Mantel-Haenszel Test) Introduction In a stratified design, the subects are selected from two or more strata which are formed from imortant

More information

A SPARSENESS CONTROLLED PROPORTIONATE ALGORITHM FOR ACOUSTIC ECHO CANCELLATION

A SPARSENESS CONTROLLED PROPORTIONATE ALGORITHM FOR ACOUSTIC ECHO CANCELLATION 6th European Signal Processing Conference (EUSIPCO 28), Lausanne, Switzerland, August 25-29, 28, copyright by EURASIP A SPARSENESS CONTROLLED PROPORTIONATE ALGORITHM FOR ACOUSTIC ECHO CANCELLATION Pradeep

More information

Research Article Estimation of Observer Parameters for Dynamic Positioning Ships

Research Article Estimation of Observer Parameters for Dynamic Positioning Ships Mathematical Problems in Engineering Volume 13, Article ID 17363, 7 ages htt://dx.doi.org/1.1155/13/17363 Research Article Estimation of Observer Parameters for Dynamic Positioning Shis Xiaogong Lin, Yehai

More information

Efficient Use Of Sparse Adaptive Filters

Efficient Use Of Sparse Adaptive Filters Efficient Use Of Sparse Adaptive Filters Andy W.H. Khong and Patrick A. Naylor Department of Electrical and Electronic Engineering, Imperial College ondon Email: {andy.khong, p.naylor}@imperial.ac.uk Abstract

More information

A Comparison between Biased and Unbiased Estimators in Ordinary Least Squares Regression

A Comparison between Biased and Unbiased Estimators in Ordinary Least Squares Regression Journal of Modern Alied Statistical Methods Volume Issue Article 7 --03 A Comarison between Biased and Unbiased Estimators in Ordinary Least Squares Regression Ghadban Khalaf King Khalid University, Saudi

More information

Robust Performance Design of PID Controllers with Inverse Multiplicative Uncertainty

Robust Performance Design of PID Controllers with Inverse Multiplicative Uncertainty American Control Conference on O'Farrell Street San Francisco CA USA June 9 - July Robust Performance Design of PID Controllers with Inverse Multilicative Uncertainty Tooran Emami John M Watkins Senior

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

Damage Identification from Power Spectrum Density Transmissibility

Damage Identification from Power Spectrum Density Transmissibility 6th Euroean Worksho on Structural Health Monitoring - h.3.d.3 More info about this article: htt://www.ndt.net/?id=14083 Damage Identification from Power Sectrum Density ransmissibility Y. ZHOU, R. PERERA

More information

Research Article Positive Solutions of Sturm-Liouville Boundary Value Problems in Presence of Upper and Lower Solutions

Research Article Positive Solutions of Sturm-Liouville Boundary Value Problems in Presence of Upper and Lower Solutions International Differential Equations Volume 11, Article ID 38394, 11 ages doi:1.1155/11/38394 Research Article Positive Solutions of Sturm-Liouville Boundary Value Problems in Presence of Uer and Lower

More information

Beamspace MIMO-NOMA for Millimeter-Wave Communications Using Lens Antenna Arrays

Beamspace MIMO-NOMA for Millimeter-Wave Communications Using Lens Antenna Arrays Beamsace MIMO-NOMA for Millimeter-Wave Communications Using Lens Antenna Arrays Bichai Wang 1, Linglong Dai 1,XiqiGao, and Lajos anzo 3 1 Tsinghua National Laboratory for Information Science and Technology

More information

The Numerical Simulation of Gas Turbine Inlet-Volute Flow Field

The Numerical Simulation of Gas Turbine Inlet-Volute Flow Field World Journal of Mechanics, 013, 3, 30-35 doi:10.436/wjm.013.3403 Published Online July 013 (htt://www.scir.org/journal/wjm) The Numerical Simulation of Gas Turbine Inlet-Volute Flow Field Tao Jiang 1,

More information

The Noise Power Ratio - Theory and ADC Testing

The Noise Power Ratio - Theory and ADC Testing The Noise Power Ratio - Theory and ADC Testing FH Irons, KJ Riley, and DM Hummels Abstract This aer develos theory behind the noise ower ratio (NPR) testing of ADCs. A mid-riser formulation is used for

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

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

Research Article A Wavelet Algorithm for Fourier-Bessel Transform Arising in Optics

Research Article A Wavelet Algorithm for Fourier-Bessel Transform Arising in Optics Hindawi Publishing Cororation International Engineering Mathematics Volume 25, Article ID 789675, 9 ages htt://dx.doi.org/.55/25/789675 Research Article A Wavelet Algorithm for Fourier-Bessel Transform

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

Computer arithmetic. Intensive Computation. Annalisa Massini 2017/2018

Computer arithmetic. Intensive Computation. Annalisa Massini 2017/2018 Comuter arithmetic Intensive Comutation Annalisa Massini 7/8 Intensive Comutation - 7/8 References Comuter Architecture - A Quantitative Aroach Hennessy Patterson Aendix J Intensive Comutation - 7/8 3

More information

On Line Parameter Estimation of Electric Systems using the Bacterial Foraging Algorithm

On Line Parameter Estimation of Electric Systems using the Bacterial Foraging Algorithm On Line Parameter Estimation of Electric Systems using the Bacterial Foraging Algorithm Gabriel Noriega, José Restreo, Víctor Guzmán, Maribel Giménez and José Aller Universidad Simón Bolívar Valle de Sartenejas,

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

Generalized Coiflets: A New Family of Orthonormal Wavelets

Generalized Coiflets: A New Family of Orthonormal Wavelets Generalized Coiflets A New Family of Orthonormal Wavelets Dong Wei, Alan C Bovik, and Brian L Evans Laboratory for Image and Video Engineering Deartment of Electrical and Comuter Engineering The University

More information

Uncorrelated Multilinear Principal Component Analysis for Unsupervised Multilinear Subspace Learning

Uncorrelated Multilinear Principal Component Analysis for Unsupervised Multilinear Subspace Learning TNN-2009-P-1186.R2 1 Uncorrelated Multilinear Princial Comonent Analysis for Unsuervised Multilinear Subsace Learning Haiing Lu, K. N. Plataniotis and A. N. Venetsanooulos The Edward S. Rogers Sr. Deartment

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

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

Design of NARMA L-2 Control of Nonlinear Inverted Pendulum

Design of NARMA L-2 Control of Nonlinear Inverted Pendulum International Research Journal of Alied and Basic Sciences 016 Available online at www.irjabs.com ISSN 51-838X / Vol, 10 (6): 679-684 Science Exlorer Publications Design of NARMA L- Control of Nonlinear

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

An Analysis of Reliable Classifiers through ROC Isometrics

An Analysis of Reliable Classifiers through ROC Isometrics An Analysis of Reliable Classifiers through ROC Isometrics Stijn Vanderlooy s.vanderlooy@cs.unimaas.nl Ida G. Srinkhuizen-Kuyer kuyer@cs.unimaas.nl Evgueni N. Smirnov smirnov@cs.unimaas.nl MICC-IKAT, Universiteit

More information

Th P10 13 Alternative Misfit Functions for FWI Applied to Surface Waves

Th P10 13 Alternative Misfit Functions for FWI Applied to Surface Waves Th P0 3 Alternative Misfit Functions for FWI Alied to Surface Waves I. Masoni* Total E&P, Joseh Fourier University), R. Brossier Joseh Fourier University), J. Virieux Joseh Fourier University) & J.L. Boelle

More information

Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling

Noncircular Sources-Based Sparse Representation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Coupling algorithms Article Noncircular Sources-Based Sarse Reresentation Algorithm for Direction of Arrival Estimation in MIMO Radar with Mutual Couling Weidong Zhou, Jing Liu *, Pengxiang Zhu, Wenhe Gong and

More information

Time Domain Calculation of Vortex Induced Vibration of Long-Span Bridges by Using a Reduced-order Modeling Technique

Time Domain Calculation of Vortex Induced Vibration of Long-Span Bridges by Using a Reduced-order Modeling Technique 2017 2nd International Conference on Industrial Aerodynamics (ICIA 2017) ISBN: 978-1-60595-481-3 Time Domain Calculation of Vortex Induced Vibration of Long-San Bridges by Using a Reduced-order Modeling

More information

On Fractional Predictive PID Controller Design Method Emmanuel Edet*. Reza Katebi.**

On Fractional Predictive PID Controller Design Method Emmanuel Edet*. Reza Katebi.** On Fractional Predictive PID Controller Design Method Emmanuel Edet*. Reza Katebi.** * echnology and Innovation Centre, Level 4, Deartment of Electronic and Electrical Engineering, University of Strathclyde,

More information

Kochi University of Technology Aca Developments of Adaptive Filter A Title parse Channel Estimation Author(s) LI, Yingsong Citation 高知工科大学, 博士論文. Date of 2014-03 issue URL http://hdl.handle.net/10173/1119

More information

Research Article A New Sum Analogous to Gauss Sums and Its Fourth Power Mean

Research Article A New Sum Analogous to Gauss Sums and Its Fourth Power Mean e Scientific World Journal, Article ID 139725, ages htt://dx.doi.org/10.1155/201/139725 Research Article A New Sum Analogous to Gauss Sums and Its Fourth Power Mean Shaofeng Ru 1 and Weneng Zhang 2 1 School

More information

Algebraic Parameter Estimation of Damped Exponentials

Algebraic Parameter Estimation of Damped Exponentials Algebraic Parameter Estimation of Damed Exonentials Aline Neves, Maria Miranda, Mamadou Mbou To cite this version: Aline Neves, Maria Miranda, Mamadou Mbou Algebraic Parameter Estimation of Damed Exonentials

More information

Indirect Rotor Field Orientation Vector Control for Induction Motor Drives in the Absence of Current Sensors

Indirect Rotor Field Orientation Vector Control for Induction Motor Drives in the Absence of Current Sensors Indirect Rotor Field Orientation Vector Control for Induction Motor Drives in the Absence of Current Sensors Z. S. WANG *, S. L. HO ** * College of Electrical Engineering, Zhejiang University, Hangzhou

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

sensors ISSN

sensors ISSN Sensors 2011, 11, 260-277; doi:10.3390/s110100260 OPEN ACCESS sensors ISSN 1424-8220 www.mdi.com/journal/sensors Article Roller Bearing Fault Diagnosis Based on Nonlinear Redundant Lifting Wavelet Pacet

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

MULTIRATE SYSTEMS AND FILTER BANKS

MULTIRATE SYSTEMS AND FILTER BANKS T. Saramäki and R. Bregovi, ultirate Systems and Filter Banks, Chater in ultirate Systems: Design T. Saramäki and R. Bregovi, ultirate Systems and Filter Banks, Chater in ultirate Systems: Design ULTIRATE

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

PARTIAL FACE RECOGNITION: A SPARSE REPRESENTATION-BASED APPROACH. Luoluo Liu, Trac D. Tran, and Sang Peter Chin

PARTIAL FACE RECOGNITION: A SPARSE REPRESENTATION-BASED APPROACH. Luoluo Liu, Trac D. Tran, and Sang Peter Chin PARTIAL FACE RECOGNITION: A SPARSE REPRESENTATION-BASED APPROACH Luoluo Liu, Trac D. Tran, and Sang Peter Chin Det. of Electrical and Comuter Engineering, Johns Hokins Univ., Baltimore, MD 21218, USA {lliu69,trac,schin11}@jhu.edu

More information

Energy-aware optimisation for run-time reconfiguration

Energy-aware optimisation for run-time reconfiguration Energy-aware otimisation for run-time reconfiguration Tobias Becker and Wayne Luk Deartment of Comuting Imerial College London, UK eter Y. K. Cheung Deartment of Electrical and Electronic Engineering Imerial

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

Improved Perfectly Matched Layers for Acoustic Radiation and Scattering Problems

Improved Perfectly Matched Layers for Acoustic Radiation and Scattering Problems NATO Undersea Research Centre Partnering for Maritime Innovation Presented at the COMSOL Conference 008 Hannover Acoustics Session Wed 5 November 008, 13:00 15:40. Imroved Perfectly Matched Layers for

More information

L P -NORM NON-NEGATIVE MATRIX FACTORIZATION AND ITS APPLICATION TO SINGING VOICE ENHANCEMENT. Tomohiko Nakamura and Hirokazu Kameoka,

L P -NORM NON-NEGATIVE MATRIX FACTORIZATION AND ITS APPLICATION TO SINGING VOICE ENHANCEMENT. Tomohiko Nakamura and Hirokazu Kameoka, L P -NORM NON-NEGATIVE MATRIX FACTORIZATION AND ITS APPLICATION TO SINGING VOICE ENHANCEMENT Tomohio Naamura and Hiroazu Kameoa, Graduate School of Information Science and Technology, The University of

More information

Hotelling s Two- Sample T 2

Hotelling s Two- Sample T 2 Chater 600 Hotelling s Two- Samle T Introduction This module calculates ower for the Hotelling s two-grou, T-squared (T) test statistic. Hotelling s T is an extension of the univariate two-samle t-test

More information

A multiplicative algorithm for convolutive non-negative matrix factorization based on squared euclidean distance

A multiplicative algorithm for convolutive non-negative matrix factorization based on squared euclidean distance Loughborough University Institutional Reository A multilicative algorithm for convolutive non-negative matrix factorization based on squared euclidean distance This item was submitted to Loughborough University's

More information

Solution of time domain electric field integral equation for arbitrarily shaped dielectric bodies using an unconditionally stable methodology

Solution of time domain electric field integral equation for arbitrarily shaped dielectric bodies using an unconditionally stable methodology RADIO SCIENCE, VOL. 38, NO. 3, 148, doi:1.129/22rs2759, 23 Solution of time domain electric field integral equation for arbitrarily shaed dielectric bodies using an unconditionally stable methodology Young-seek

More information

Research of PMU Optimal Placement in Power Systems

Research of PMU Optimal Placement in Power Systems Proceedings of the 5th WSEAS/IASME Int. Conf. on SYSTEMS THEORY and SCIENTIFIC COMPUTATION, Malta, Setember 15-17, 2005 (38-43) Research of PMU Otimal Placement in Power Systems TIAN-TIAN CAI, QIAN AI

More information

COMPLEX-VALUED ASSOCIATIVE MEMORIES WITH PROJECTION AND ITERATIVE LEARNING RULES

COMPLEX-VALUED ASSOCIATIVE MEMORIES WITH PROJECTION AND ITERATIVE LEARNING RULES JAISCR, 2018, Vol 8, o 3, 237 249 101515/jaiscr-2018-0015 COMPLEX-VALUED ASSOCIATIVE MEMORIES WITH PROJECTIO AD ITERATIVE LEARIG RULES Teijiro Isokawa 1, Hiroki Yamamoto 1, Haruhiko ishimura 2, Takayuki

More information

Proximal methods for the latent group lasso penalty

Proximal methods for the latent group lasso penalty Proximal methods for the latent grou lasso enalty The MIT Faculty has made this article oenly available. Please share how this access benefits you. Your story matters. Citation As Published Publisher Villa,

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

Observer/Kalman Filter Time Varying System Identification

Observer/Kalman Filter Time Varying System Identification Observer/Kalman Filter Time Varying System Identification Manoranjan Majji Texas A&M University, College Station, Texas, USA Jer-Nan Juang 2 National Cheng Kung University, Tainan, Taiwan and John L. Junins

More information

MULTI-CHANNEL PARAMETRIC ESTIMATOR FAST BLOCK MATRIX INVERSES

MULTI-CHANNEL PARAMETRIC ESTIMATOR FAST BLOCK MATRIX INVERSES MULTI-CANNEL ARAMETRIC ESTIMATOR FAST BLOCK MATRIX INVERSES S Lawrence Marle Jr School of Electrical Engineering and Comuter Science Oregon State University Corvallis, OR 97331 Marle@eecsoregonstateedu

More information

Learning Sequence Motif Models Using Gibbs Sampling

Learning Sequence Motif Models Using Gibbs Sampling Learning Sequence Motif Models Using Gibbs Samling BMI/CS 776 www.biostat.wisc.edu/bmi776/ Sring 2018 Anthony Gitter gitter@biostat.wisc.edu These slides excluding third-arty material are licensed under

More information

Information collection on a graph

Information collection on a graph Information collection on a grah Ilya O. Ryzhov Warren Powell February 10, 2010 Abstract We derive a knowledge gradient olicy for an otimal learning roblem on a grah, in which we use sequential measurements

More information

Rotations in Curved Trajectories for Unconstrained Minimization

Rotations in Curved Trajectories for Unconstrained Minimization Rotations in Curved rajectories for Unconstrained Minimization Alberto J Jimenez Mathematics Deartment, California Polytechnic University, San Luis Obiso, CA, USA 9407 Abstract Curved rajectories Algorithm

More information

Dynamic System Eigenvalue Extraction using a Linear Echo State Network for Small-Signal Stability Analysis a Novel Application

Dynamic System Eigenvalue Extraction using a Linear Echo State Network for Small-Signal Stability Analysis a Novel Application Dynamic System Eigenvalue Extraction using a Linear Echo State Network for Small-Signal Stability Analysis a Novel Alication Jiaqi Liang, Jing Dai, Ganesh K. Venayagamoorthy, and Ronald G. Harley Abstract

More information

Evaluating Circuit Reliability Under Probabilistic Gate-Level Fault Models

Evaluating Circuit Reliability Under Probabilistic Gate-Level Fault Models Evaluating Circuit Reliability Under Probabilistic Gate-Level Fault Models Ketan N. Patel, Igor L. Markov and John P. Hayes University of Michigan, Ann Arbor 48109-2122 {knatel,imarkov,jhayes}@eecs.umich.edu

More information

Position Control of Induction Motors by Exact Feedback Linearization *

Position Control of Induction Motors by Exact Feedback Linearization * BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 8 No Sofia 008 Position Control of Induction Motors by Exact Feedback Linearization * Kostadin Kostov Stanislav Enev Farhat

More information

OPTIMISATION OF TRANSMISSION PREDICTIONS FOR A SONAR PERFORMANCE MODEL FOR SHALLOW OCEAN REGIONS

OPTIMISATION OF TRANSMISSION PREDICTIONS FOR A SONAR PERFORMANCE MODEL FOR SHALLOW OCEAN REGIONS OPTIMISATION OF TRANSMISSION PREDICTIONS FOR A SONAR PERFORMANCE MODEL FOR SHALLOW OCEAN REGIONS Adrian D. Jones*, Janice S. Sendt, Z. Yong Zhang*, Paul A. Clarke* and Jarrad R. Exelby* *Maritime Oerations

More information

x and y suer from two tyes of additive noise [], [3] Uncertainties e x, e y, where the only rior knowledge is their boundedness and zero mean Gaussian

x and y suer from two tyes of additive noise [], [3] Uncertainties e x, e y, where the only rior knowledge is their boundedness and zero mean Gaussian A New Estimator for Mixed Stochastic and Set Theoretic Uncertainty Models Alied to Mobile Robot Localization Uwe D. Hanebeck Joachim Horn Institute of Automatic Control Engineering Siemens AG, Cororate

More information

Optimal Recognition Algorithm for Cameras of Lasers Evanescent

Optimal Recognition Algorithm for Cameras of Lasers Evanescent Otimal Recognition Algorithm for Cameras of Lasers Evanescent T. Gaudo * Abstract An algorithm based on the Bayesian aroach to detect and recognise off-axis ulse laser beams roagating in the atmoshere

More information

Uniform Law on the Unit Sphere of a Banach Space

Uniform Law on the Unit Sphere of a Banach Space Uniform Law on the Unit Shere of a Banach Sace by Bernard Beauzamy Société de Calcul Mathématique SA Faubourg Saint Honoré 75008 Paris France Setember 008 Abstract We investigate the construction of a

More information

Keywords: Vocal Tract; Lattice model; Reflection coefficients; Linear Prediction; Levinson algorithm.

Keywords: Vocal Tract; Lattice model; Reflection coefficients; Linear Prediction; Levinson algorithm. Volume 3, Issue 6, June 213 ISSN: 2277 128X International Journal of Advanced Research in Comuter Science and Software Engineering Research Paer Available online at: www.ijarcsse.com Lattice Filter Model

More information

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

Numerical Linear Algebra

Numerical Linear Algebra Numerical Linear Algebra Numerous alications in statistics, articularly in the fitting of linear models. Notation and conventions: Elements of a matrix A are denoted by a ij, where i indexes the rows and

More information

Adaptive Minimum Symbol-Error-Rate CDMA Linear Multiuser Detector for Pulse-Amplitude Modulation

Adaptive Minimum Symbol-Error-Rate CDMA Linear Multiuser Detector for Pulse-Amplitude Modulation Adative Minimum Symbol-Error-Rate CDMA inear Multiuser Detector for Pulse-Amlitude Modulation AK Samingan, S Chen and Hanzo Deartment of Electronics and Comuter Science University of Southamton, Southamton

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

ADAPTIVE CONTROL METHODS FOR EXCITED SYSTEMS

ADAPTIVE CONTROL METHODS FOR EXCITED SYSTEMS ADAPTIVE CONTROL METHODS FOR NON-LINEAR SELF-EXCI EXCITED SYSTEMS by Michael A. Vaudrey Dissertation submitted to the Faculty of the Virginia Polytechnic Institute and State University in artial fulfillment

More information

Frequency-Weighted Robust Fault Reconstruction Using a Sliding Mode Observer

Frequency-Weighted Robust Fault Reconstruction Using a Sliding Mode Observer Frequency-Weighted Robust Fault Reconstruction Using a Sliding Mode Observer C.P. an + F. Crusca # M. Aldeen * + School of Engineering, Monash University Malaysia, 2 Jalan Kolej, Bandar Sunway, 4650 Petaling,

More information

Seafloor Reflectivity A Test of an Inversion Technique

Seafloor Reflectivity A Test of an Inversion Technique Seafloor Reflectivity A Test of an Inversion Technique Adrian D. Jones 1, Justin Hoffman and Paul A. Clarke 1 1 Defence Science and Technology Organisation, Australia, Student at Centre for Marine Science

More information