Designing and predicting QoS of a wireless system for medical telemetry

Guaranteeing high reliability and acceptable latency are major QoS considerations in the design of life-critical healthcare applications. To achieve those goals over error-prone wireless networks, proper error control is required. We propose a reference system model for wireless telemetry with medical-grade QoS. The system includes the combination of delay-predictable medium access control and interleaved forward error correction (FEC) based on Reed-Solomon (RS) coding. This combination is effective in correcting error bursts and bounding the service latency. However, variation in the delay incurred by RS decoding can lead to service dropout. We therefore determine the size of buffer required to absorb this jitter, and to stochastically guarantee seamless services over a fading channel, using a worst-case execution time analysis. But this buffering itself leads to an increase in service latency, which grows with the level of block interleaving. The QoS analysis tool which we have developed investigates this tradeoff, and hence it is effective in finding service parameters that simultaneously satisfy the requirements of reliability and latency in medical applications.