In CDMA systems, outer loop power control (OLPC) determines the target value of SNR at the receiver, mostly by using look-up tables to map bit error rates (BERs) to SNR-targets. In this contribution, transmission delay and packet loss rate constraints in the data link layer (DLL) are invoked in order to determine the optimum outer loop SNR-target setpoint analytically, according to the number of active users in cell. Optimality is, in this sense, the maximization of system throughput. Using the optimum SNR-target, the optimal spreading factor is determined. Subsequently, the joint optimization of outer loop SNR-target and variable spreading factor (VSF), at the physical(PHY)-layer, with truncated automatic repeat request (ARQ) error control mechanism at the data link layer are proposed. Hence, we show that quality of service (QoS) requirements at these layers can be simultaneously satisfied while maximizing throughput. Total and truncated channel inversion strategies are employed in the inner loop to adapt transmit power to short-time channel variations. We propose a system where the number of users in a cell is modeled by a one-dimensional discrete Markov chain, and design the adaptive continuous power and rate mechanism for the worst case packet error rate (PER) condition. The corresponding theoretical throughput, which can be regarded as upper-bound for discrete spreading factor case, is obtained numerically for various settings of system parameters. We have also provided simulation results for a practical channel condition. Our scheme is compared with “constant SNR-target” and “PHY-layer based variable SNR-target” cases under continuous power and rate variation to show the achievable gain through the coupling of physical and data link layers parameters.
Published in: IEEE Transactions on Wireless Communications ( Volume: 10, Issue: 10, October 2011)
Page(s): 3278 – 3287
Date of Publication: 18 August 2011
Introduction
Effective and efficient resource allocation strategies are vital in achieving desirable performance in direct sequence code division multiple access (DS-CDMA) systems. The significance of this concept is heightened by the need for accommodating multiple classes of traffic (e.g. voice, data, and compressed video) with different quality of service (QoS) requirements. In particular, QoS provisioning in 3rd generation networks with Femto-cell layers requires more efficient techniques for allocation of limited resources [1], [2]. Adaptive resource allocation methods have been widely considered in the last decade. These schemes exploit the time-varying nature of a channel in order to manage resources optimally, yielding better throughput. Typical adaptive techniques include adaptation of transmitter power [3]–[5], adaptive modulation [6], [7], adaptive coding [8], adaptive spreading factor [9]–[17], or any combination of these methods [10], [11], [18], [19].
Read More : https://ieeexplore.ieee.org/document/5989893
