Shannon Wireless’ Post

This paper leverages model pruning and proposes a #pruning-#enabled #hierarchical #federated #learning (#PHFL) in #heterogeneous #networks (#HetNets). The authors first derive an upper bound of the convergence rate that clearly demonstrates the impact of the model pruning and wireless communications between the clients and the associated BS. Then they jointly optimize the model pruning ratio, #central #processing #unit (#CPU) frequency and transmission power of the clients in order to minimize the controllable terms of the convergence bound under strict delay and energy constraints. However, since the original problem is not convex, they perform #successive #convex #approximation (#SCA) and jointly optimize the parameters for the relaxed convex problem. ---- Ferdous Pervej, Richeng Jin, Huaiyu Dai More details can be found at this link: https://lnkd.in/eveBgZB8

This letter introduced a novel #optical #DACs (#oDAC) parallelizing a pair of optimized 2-segment MZMs, remarkably fed by 4 uncoupled NRZ signals. Based on a pair of IQ-nested new oDACs, the resulting DAC-less QAM #Transmitters (#Tx) is fed by 8 uncoupled NRZ drivers. The energy efficiency, photonic efficiency and resolution of the flexible oDAC and QAM Tx are at their utmost. ---- Moshe Nazarathy, Dr. Ioannis Tomkos More details can be found at this link: https://lnkd.in/eBQJs8KZ

In this article, the authors show that #Reservoir #Computing (#RC) can universally approximate a general #linear #time-#invariant (#LTI) system. Specifically, they present a clear signal processing interpretation of RC and utilize this understanding in the problem of approximating a generic LTI system. Under this setup, they analytically characterize the optimum probability density function for configuring (instead of training and/or randomly generating) the recurrent weights of the underlying #recurrent #neural #network (#RNN) of the RC. ---- Shashank Jere, Lizhong Zheng, Karim Said, Lingjia Liu More details can be found at this link: https://lnkd.in/eVnTTxVH

In this paper, the application of #integrated #access and #backhaul (#IAB) technique in #millimeter-#wave (#mmWave) #cell-#free (#CF) massive MIMO systems is investigated. The access and backhaul links share a frequency spectrum in the mmWave bands, and in both, hybrid beamforming techniques are adopted for signal transmission. The bandwidth allocation (division) parameter between the two link types as well as the beamforming matrices are optimized to maximize the end-to-end data rate. This leads to a non-convex optimization problem for which an efficient solution method is proposed. ---- Ali Hosseinalipour jazi, S. Mohammad Razavizadeh, Tommy Svensson More details can be found at this link: https://lnkd.in/eB_q8BTJ

This letter design a low-complexity #minimum #mean #square #error (#MMSE) receiver for a practical rectangular pulse-shaped #orthogonal #time #frequency #space (#OTFS) #massive #multiple-#input #multiple-#output (#mMIMO) system. The proposed receiver is designed by exploiting the OTFS channel sparsity, and the structure of matrices involved in the MMSE receiver. The authors show that the sparse OTFS channel matrix is multi-banded, and has a high bandwidth. This is because its non-zero elements, which occur in multiple bands, are scattered throughout the matrix. They reduce its bandwidth by using the Reverse Cuthill-Mckee algorithm. The proposed MMSE receiver does not use any approximation, and thus has the same #bit #error #rate (#BER) as that of its conventional counterpart. ---- Mudasir Ahmad sheikh, Prem Singh, Rohit Budhiraja More details can be found at this link: https://lnkd.in/ezpXv7S4

This article presents the designs of ultrawideband microwave flat #gradient #index (#GRIN) #lenses, which realizes over a 108% fractional bandwidth (12-40 GHz). The frequency-independent ray optics method is employed to determine the radially varying permittivity profile of the lenses. The challenge of realizing such a radially varying profile and the limitations in dielectric material choices are overcome by two additive-manufacturing-aided approaches: 1) partially infilled dielectrics with a varied periodicity, which ensures the lens performance at the higher end of the frequency range and 2) #artificially #engineered #dielectrics (#AED) with subwavelength-scale metallic inclusions, which enables-high permittivity dielectrics and leads to benefits of thickness and mass reduction for the GRIN lenses. ---- Shiyu Zhang, Ravi Kumar ARYA ,Prof. Will Whittow, Darren Cadman, Raj Mittra, J. C. Vardaxoglou More details can be found at this link: https://lnkd.in/eaHHDtV8

This article investigates the double-#reconfigurable #intelligent #surfaces (#RISs) channel characteristics and propose a #geometry-#based #triple-#cylinder #model, where the RIS subsurface/tile is enabled to assist #vehicle-#to-#vehicle (#V2V) systems and other elements are turned off. To determine tile locations on RIS surface, they formulate an optimization problem by maximizing the end-to-end channel gain and solve it using the #gradient #ascent (#GA) method. Following this, channel correlation function, channel capacity, and outage probability are derived according to the proposed model. Five typical mobile scenarios are discussed to validate the convergence of proposed GA algorithm, where the results show that channel gain can converge to its maximum with optimized tile locations.----@Guiqi Sun, Ruisi He, Jiancheng An, @Bo Ai, @Yaxin Song, @Yong Niu, @Gongpu Wang, Chau Yuen More details can be found at this link: https://lnkd.in/gMskqEPg

This paper proposes a scheme called #joint #delay-#Doppler #index #modulation #orthogonal #time #frequency #space (#JDDIM-#OTFS) to enhance reliability in OTFS systems to satisfy the rigorous requirements of forthcoming communication systems. This scheme activates delay or #Doppler #resource elements in a subframe, and in order to obtain higher spectral efficiency, the authors use different constellations for each active resource block. Furthermore, they utilize a #low-#complex #detector with the help of a #greedy #approach. Moreover, a theoretical upper bound is derived for the #bit #error #rate (#BER). ----Yusuf İslam Tek, Ertugrul Basar More details can be found at this link: https://lnkd.in/grByheMG

This paper proposes an #over-#the-#air (#OTA)-based approach for distributed matrix-vector multiplications in the context of #distributed #machine #learning (#DML). Thanks to OTA computation, the column-wise partitioning of a large matrix enables efficient workload distribution among workers (i.e., #local #computing #nodes) based on their computing capabilities. In addition, without requiring additional bandwidth, it allows the system to remain scalable even as the number of workers increases to mitigate the impact of slow workers, known as stragglers. However, despite the improvements, there are still instances where some workers experience deep fading and become stragglers, preventing them from transmitting their results. By analyzing the #mean #squared #error (#MSE), they demonstrate that incorporating more workers in the OTA-based approach leads to MSE reduction without the need for additional radio resources. ----Jinho Choi More details can be found at this link: https://lnkd.in/gm4uGWMi