Multi-dimensional direct-sequence spread spectrum multiple-access communication with adaptive channel coding

Abstract

During the race towards the4th generation (4G) cellular-based digital communication systems, a growth in the demand for high capacity, multi-media capable, improved Quality-of-Service (QoS) mobile communication systems have caused the developing mobile communications world to turn towards betterMultiple Access (MA) techniques, like Code Division Multiple Access (CDMA) [5]. The demand for higher throughput and better QoS in future 4G systems have also given rise to a scheme that is becoming ever more popular for use in these so-called ‘bandwidth-on-demand’ systems. This scheme is known as adaptive channel coding, and gives a system the ability to firstly sense changes in conditions, and secondly, to adapt to these changes, exploiting the fact that under good channel conditions, a very simple or even no channel coding scheme can be used for Forward Error Correction(FEC). This will ultimately result in better system throughput utilization. One such scheme, known as incremental redundancy, is already implemented in the Enhanced Data Rates for GSM Evolution (EDGE) standard. This study presents an extensive simulation study of a Multi-User (MU), adaptive channel coded Direct Sequence Spread Spectrum Multiple Access (DS/SSMA) communication system. This study firstly presents and utilizes a complex Base Band(BB) DS/SSMA transmitter model, aimed at user data diversity [6] in order to realize the MU input data to the system. This transmitter employs sophisticated double-sideband (DSB)Constant-Envelope Linearly Interpolated Root-of-Unity (CE-LI-RU) filtered General Chirp-Like (GCL) sequences [34, 37, 38] to band limit and spread user data. It then utilizes a fully user-definable, complex Multipath Fading Channel Simulator(MFCS), first presented by Staphorst [3], which is capable of reproducing all of the physical attributes of realistic mobile fading channels. Next, this study presents a matching DS/SSMA receiver structure that aims to optimally recover user data from the channel, ensuring the achievement of data diversity. In order to provide the basic channel coding functionality needed by the system of this study, three simple, but well-known channel coding schemes are investigated and employed. These are: binary Hamming (7,4,3) block code, (15,7,5) binary Bose-Chadhuri-Hocquenghem (BCH) block code and a rate 1/3 <i.Non-Systematic (NS) binary convolutional code [6]. The first step towards the realization of any adaptive channel coded system is the ability to measure channel conditions as fast as possible, without the loss of accuracy or inclusion of known data. In 1965, Gooding presented a paper in which he described a technique that measures communication conditions at the receiving end of a system through a device called a Performance Monitoring Unit (PMU) [12, 13]. This device accelerates the system’sBit Error Rate (BER) to a so-called Pseudo Error Rate(PER) through a process known as threshold modification. It then uses a simple PER extrapolation algorithm to estimate the system’s true BER with moderate accuracy and without the need for known data. This study extends the work of Gooding by applying his technique to the DS/SSMA system that utilizes a generic Soft-Output Viterbi Algorithm(SOVA) decoder [39] structure for the trellis decoding of the binary linear block codes [3, 41-50], as well as binary convolutional codes mentioned, over realistic MU frequency selective channel conditions. This application will grant the system the ability to sense changes in communication conditions through real-time BER measurement and, ultimately, to adapt to these changes by switching to different channel codes. Because no previous literature exists on this application, this work is considered novel. Extensive simulation results also investigate the linearity of the PER vs. modified threshold relationship for uncoded, as well as all coded cases. These simulations are all done for single, as well as multiple user systems. This study also provides extensive simulation results that investigate the calculation accuracy and speed advantages that Gooding’s technique possesses over that of the classic Monte-Carlo technique for BER estimation. These simulations also consider uncoded and coded cases, as well as single and multiple users. Finally, this study investigates the experimental real-time performance of the fully functional MU, adaptive coded, DS/SSMA communication system over varying channel conditions. During this part of the study, the channel conditions are varied over time, and the system’s adaptation (channel code switching) performance is observed through a real-time observation of the system’s estimated BER. This study also extends into cases with multiple system users. Since the adaptive coded system of this study does not require known data sequences (training sequences), inclusion of Gooding’s technique for real-time BER estimation through threshold modification and PER extrapolation in future 4G adaptive systems will enable better Quality-of-Service (QoS) management without sacrificing throughput. Furthermore, this study proves that when Gooding’s technique is applied to a coded system with a soft-output, it can be an effective technique for QoS monitoring, and should be considered in 4G systems of the future.