In order to understand the complete behavior of the complex indoor or outdoor propagation environment, direct channel measurement is required. This area of research requires the development of hardware and software to measure indoor and outdoor wireless channels. This data also serves as an input for development of accurate and reliable channel models.
Characterization of the performance of advanced antenna systems requires an understanding of the joint spatial and temporal response of the channel. Our early efforts in channel probing focused on measuring the temporal channel response indoors with very high temporal resolution (~2 ns). Utilizing a directional dish antenna at the receiver, we also can obtain the angular resoponse of the channel.
Recently, we have developed and built a flexible switched-architecture wideband channel sounder, supporting about 100 MHz of instantaneous bandwidth and up to eight transmit and receive antennas. We are using data collected with this platform to study the effects of time-variation in MIMO channels. We intend to share details of this system with other researchers who are interested. This platform has also been built at the University of Pretoria in South Africa.
Narrowband MIMO Channel Probing
Previous channel measurement efforts have focused on the development of a measurement platform for the narrowband MIMO (multiple-input multiple-output) wireless channel. We have developed a system with up to 16 transmit and 16 receive antennas. The system can be easily reconfigured, allwoing us to probe multiple bands.
Through our measurement efforts, we have discovered a number of interesting facts about many aspects of MIMO wireless communications. For example, we have learned the influence of antenna polarization, spacing, and directivity on channel capacity capacity.
Since measured channel data is often limited in scope and difficult to obtain, accurate and reliable channel models are highly desirable. Channel models range from very approximate statistical models to advanced ray-tracing techniques.
One of our areas of research has been the extension of the physical path-based Saleh-Valenzuela model to include spatial (angular) information. Combination of the path-based model with detailed electromagnetic antenna modeling to a hybrid model for assessing the performance MIMO communications systems.
Initial studies of space-time coding essentially assumed a constant channel. However, when the channel is time-varying, estimates of the channel obtained by training are often very poor and quickly lead to algorithmic degradation. We have studied the effect of a time-varying channel on both trained and differential space-time coding techniques, showing that time-varyiation can often be treated as an effective degradation of the SNR, and that differential coding can often outperform trained schemes.
Measured channel data provides real-world channels for
algorithmic assessment, but measured data is limited and may be hard
to obtain. Advanced MIMO testbeds not only exercise algorithms in
many different environments, but also assess the feasibility of
real-time implementation.
Real-time MIMO Testbed
We have developed a MIMO prototyping testbed, employing 4 transmit and receive elements (scalable to 8 elements). The architecture is based on a Pentek DSP platform, allowing rapid development of algorithms in C or C++. Embedded PCs simplify the usability of system.
Initial studies have demonstrated that employing multiple antennas at transmit and receive increase the capacity for single-user communication in the presence of multipath fading. However, real-world systems will generally serve many users. We are interested in the development of multi-antenna element architectures for the multi-user case that offer similar capacity improvements to the single-user case.
Space-Time Coding
MIMO wireless communications links
have been shown to have the potential for significant increases in
capacity, provided they are deployed in an environment with rich
multipath scattering. To realize these gains, a number of space-time
coding strategies have been proposed. We are interested in the
analysis of existing schemes, and the development of new algorithms
that address their short-comings.
After initial information theoretic studies on MIMO, many space-time coding strategies were proposed. However, little work had been done to develop a common framework in order to understand the relationship among these competing algorithms. We have developed a generalized space-time block-coding framework that treats many existing coding schemes as special cases. A diagonal space-time coding scheme leads a subspace-based algorithm that produces closed-form estimates of the channel and transmitted symbols. The algorithm is shown to be applicable to cases involving fewer receive than transmit antennas, rank-deficient channels, flat or frequency selective fading, and multiple users.
Information theory provides an upper bound on the acheivable transmission rate in communications systems, thus providing an invaluable tool for understanding the theoretical limits of communication.
Initial MIMO studies focused on the channel capacity for a fixed antenna array and ideal antennas. However, this capacity is not really the capacity of the physical channel, since the antenna array is part of the communications system that can be changed. We have devloped the notion of intrinsic capacity that defines the capacity of the physical channel, regardless of what type of communication system is employed.
Most studies on MIMO wireless communications have ignored the effect of mutual-coupling or have treated it only simplistically. A rigorous network analysis of the problem demonstrates that for small numbers of antennas, mutual coupling can indeed increase capacity. We have also found through application of an advanced noise model that capacity is usually maximized by choosing a network that minimizes amplifier noise figure, rather than providing maximum power transfer.