Multi-Antenna Testbeds for Wireless Communications
Published in George Tsoulos, MIMO System Technology for Wireless Communications, 2018
Raghu Rao, Christian Oberli, Stephan Lang, David Browne, Weijun Zhu, Mike Fitz, Babak Daneshrad
Representation of design tradeoffs: In general, there are two purposes for a testbed: (1) being a research platform for understanding wireless channels and modulations, and (2) being a technology prototype for understanding the issues of building a particular wireless application. The testbed’s architecture should match the goal. A research platform should have high performance components so that the resulting performance is a function of the channel and of the proposed algorithms and not a function of the hardware implementation. For instance, the authors’ experience recommends that a research testbed should use a digital IF, Nyquist sampling, and a digital down-converter chip. While this architecture uses significant power, it does not have I/Q mismatch as in a direct down-conversion receiver (as explained later in Section 10.4). Alternatively a technology prototype for a low cost commercial application should have a much less capable radio system, so that the algorithm development and testing can be done with realistic impairments.