Designing coding and modulation schemes for body-area networks

Abstract

A major role in the realm of health care is played by wireless monitoring systems. Body-area networks (BANs), which use the human body to support communication using low-power wireless sensor network technology, of late have been attracting a considerable interest. Now, the design of a transmission system for reliable communication through a BAN is a challenging problem. First, communication takes place on different types of links, depending on the body parts to which transmit and receive antennas are attached, e.g., trunk-to-trunk, trunk-to-head, trunk-to-hand, on where the hardware is located (body-to-body, offbody, on-body, and in-body links), and on antenna type and orientation, body size, location, and posture. In addition, propagation in on-body links may combine surface wave, creeping wave, diffracted waves, scattered waves, and free space propagation, depending on the antenna positions and the body postures. The use of multiple-input, multiple-output (MIMO) systems in BANs has also been advocated, and MIMO BAN channel models discussed. Even if adaptive techniques are used to adjust modulation and coding to the changing environment, reliable mathematical models for the transmission channel are called for, but they are difficult to obtain because of the variations of the environment in which the transmission is taking place. We argue that the design of modulation and coding schemes in BANs should be based on their robustness to uncertainties of channel model. To avoid catastrophic performance degradations due to model uncertainties, system performance must be examined by evaluating the effects of a discrepancy between the nominal distribution of channel statistics and the actual distribution. Based on this concept, the robustness of system design to channel modeling can be assessed. In this paper, we examine mathematical tools allowing designers to assess a BAN system performance under modeling uncertainty: moment-bound techniques and optimization techniques are applied to obtain performance bounds with and without error-control coding. Based on these bounds, we propose robust coding/modulation techniques making that performance less sensitive to modeling errors. Finally, we examine the design of an architecture to operate BANs in a cloud environment. Our architecture facilitates smooth convergence and operation of BANs in cloud computing, and helps the users to rely upon an efficient, reliable, and fault-tolerant cloud infrastructure for communication.