Capacity-Outage Joint Analysis and Optimal Power Allocation for Wireless Body Area Networks

Wireless body area networks (WBANs) are characterized by short-range wireless communication between inherently resource-limited sensors that operate in the vicinity of the human body for recording certain physiological signals. One of the main limitations of WBANs is in terms of their available power. Since the nodes are typically battery-driven, an efficient power transmission is essential to guarantee long-lasting communications among nodes without the need for frequent battery replacement or charging. In this study, aiming to improve the WBAN performance, we first investigate the power allocation problem by choosing the ergodic capacity and the outage probability as the desired metrics. Subject to an average power constraint, it can be shown that the power allocation, which is optimal for the ergodic capacity, does not necessarily yield a good result for the outage probability, and vice versa. Therefore, we assert that although both of the aforementioned metrics are recognized as the key performance indicators, the WBAN performance would not be appropriately characterized by considering each of them separately. Consequently, a novel hybrid metric is proposed that jointly accounts for the capacity and the outage. Subject to an average power constraint, we then find the power allocation policy that maximizes the proposed hybrid metric. The entire devised framework is utilized in the context of WBAN channels derived from applying the well-known Bayesian information criterion to a set of publicly available experimental channel data. The numerical results demonstrate the superiority of the proposed power allocation policy compared to the legacy water-filling and truncated-inversion approaches.