An energy-efficient virtual MIMO communications architecture based on V-blast processing for distributed wireless sensor networks

Abstract

An energy-efficient virtual multiple-input multiple-output (MIMO) communication architecture based on V-BLAST receiver processing is proposed for energy-constrained, distributed wireless sensor networks. The proposed scheme does not require transmitter-side sensor cooperation unlike previously proposed virtual MlMO schemes for wireless sensor networks. In sensor networks with single-antenna data gathering nodes, the virtual MIMO operation is realized via the receiver-side local communication assuming node cooperation. Numerical results show that the significant energy savings are offered by the proposed virtual MIMO architecture in distributed wireless sensor networks. These results also indicate that while rate optimization over transmission distance may offer improved energy efficiencies in some cases, this is not essential in achieving energy savings as opposed to previously proposed Alamouti scheme-based virtual MIMO implementations. In fact, in most scenarios a fixed-rate virtual MIMO system with binary phase-shift-keying (BPSK) can achieve performance very close to that of a variable-rate system with optimized rates. However, these results also indicate that the proposed scheme can lead to larger delay penalties compared to a traditional SISO communication based sensor network as the order of the virtual MIMO architecture grows. This results in a trade-off between the achievable energy efficiency and the delay incurred, making the proposed virtual V-BLAST based MIMO scheme an especially good candidate communication architecture for energy-starved and delay-tolerant wireless sensor networks having no inter-sensor communication.