Iterative Detection for Multi-User MIMO Systems

Multi-input multi-output (MIMO) systems can serve as the building blocks for spectrally efficient mobile multi-user tactical wireless systems; however, care must be taken to translate high per-link spectral efficiency into high network throughput. Mobile tactical networks generally have several features that complicate achievement of high multi-user MIMO spectral efficiency, including non-centralized, infrastructure-free operation and operation when all nodes are mobile, as well as requirements to minimize probability of interception and susceptibility to jamming. The paper presents results from both simulations and an extensive over-the-air measurement campaign, which illustrate the key features needed for spectrally efficient multi-user MIMO systems. High MIMO network throughput can be achieved by combining efficient MAC (taking into account MIMO signal processing resources) and transmitter stream control, as well as high performance MIMO signal processing and waveforms that minimize per-link signal-to-interference-and-noise ratio (SINR) requirements leading to effective spectral reuse. Iterative detection schemes provide such high spectral efficiency at low SINRs, achieving performance that approaches the Shannon bound, and provide scalable complexity. These methods are also well suited to operation in multi-user environments, supporting both interference- mitigation and joint detection approaches. The paper provides analysis and experimental results for waveforms and receiver signal processing for Space-Time Bit Interleaved Coded Modulation using multi- carrier waveforms. Results and parameter trades are presented for variations of this approach using List Sphere Detection and Soft-Symbol Cancellation as well as structures based on convolutional codes and turbo codes. See also ADM001741. Presented at the Annual Adaptive Sensor Array Processing Workshop (12th) held in Lexington, MA on 16-18 March 2004. ASAP-12-VOL-1. Briefing charts only. The original document contains color images.