Throughput-optimal cross-layer resource allocation in DS-CDMA systems with Nakagami multipath fading

Joint optimization of outer loop power control (OLPC) signal to noise ratio-target (SNR-target) and variable spreading factor (VSF), at the physical(PHY)-layer, with truncated automatic repeat request (ARQ), at data link layer (DLL), in accordance to the number of active users in the cell is considered. This investigation is on a single cell conventional cellular DS-CDMA communication system with frequency-selective fading channels, where the number of active users is modelled through a one-dimensional discrete Markov chain. The optimally is in the sense of maximizing the sum-throughput, given a coherent RAKE receiver is employed with maximum ratio combining (MRC). We determine the optimal spreading factor using OLPC SNR-target at the PHY-layer, which satisfies the QoS imposed by packet error rate-target (PER-target) as a function of maximum number of allowed ARQ retransmissions at the DLL. The channel model considers independent paths with Nakagami fading characteristics. Total and truncated channel inversion techniques are used within the inner loop power control (ILPC) to adapt the transmission power to short time channel variations. Sum-throughput performance of the optimized system and non optimized system, where the SNR-target is assumed to be constant, under various multipath fading conditions are studied. A considerable gain in sum-throughput is achieved through coupling of PHY-layer and DLL parameters.

Date of Conference: 11-14 September 2011

Date Added to IEEE Xplore26 January 2012

Conference Location: Toronto, ON, Canada


Introduction

In CDMA systems, adaptive allocation of resources in order to benefit from favourable channel conditions is essential for achieving desirable performance. Adaptive resource allocation techniques typically consist of varying the transmit power [1], modulation level [2], adaptive spreading factor [3], or any combination of these schemes [4], [5]. In Universal Mobile Telecommunications System (UMTS) standards, the uplink power control within CDMA systems consist of OLPC which operates at the frame level (duration of one frame is 10 ms), and ILPC which operates at slot level (there are 15 slots per frame) [6]. The OLPC determines the SNR-target at the receiver based on the required transmission quality mostly defined in terms of bit error rate (BER) or frame error rate (FER), and the more rapid ILPC aims to attain the SNR-target while satisfying the quality of service constraints. The ILPC achieves this by comparing the estimated received SNR, to the SNR-target and varying the transmit power (and probably spreading factor) accordingly. In practice, to accommodate for the quality requirements of users, predefined lookup tables select the OLPC SNR-target based on the received BER or FER. However, this does not achieve the optimal throughput. A number of works have analysed the improvements achieved by optimizing OLPC [7]. In particular, optimization of OLPC SNR target and variable spreading factor (VSF) by considering physical layer was studied in our previous work [8].

Read More : https://ieeexplore.ieee.org/document/6139758/

farhad zarringhalam

Farhad Zarringhalam

Biography

Farhad Zarringhalam received a BEng degree in Electronics and Ph.D. in Wireless Communications from King’s College London, in 2003 and 2007, respectively. During his PhD he worked with Nokia, UK, on adaptive methods of resource allocation and optimisation in cellular networks and published several journal and conference papers. He currently works for Quod Financial, London.