Impact of polling on Bluetooth piconet performance

Abstract

Bluetooth has become a ubiquitous technology present in almost every electronic device. A question often asked by manufacturers and final users is whether it can be used for uses other than the ones for which it was designed; in particular, for improved multimedia traffic support, with the fewest modifications in the implementation of the protocol (both hardware and software). In this article we analyze the impact on performance of the lowest levels of the Bluetooth architecture through a relevant parameter known as the polling time, Tpoll. Here, we report on simulations of specific algorithms that allow to dynamically adapt the network to an optimal value of Tpoll during transmission. We place special emphasis on the role of the kind of traffic on the network (e.g., constant bit rate, burst, multimedia-like). In addition, we analyze the results of simulations by comparing these algorithms with the static allocation value defined in the Bluetooth specification. Our results provide insight on how to optimize the transmission of multimedia traffic over piconets at the lowest layers of the Bluetooth architecture.

Bluetooth has become a ubiquitous technology present in almost every electronic device. A question often asked by manufacturers and final users is whether it can be used for uses other than the ones for which it was designed; in particular, for improved multimedia traffic support, with the fewest modifications in the implementation of the protocol (both hardware and software). In this article we analyze the impact on performance of the lowest levels of the Bluetooth architecture through a relevant parameter known as the polling time, Tpoll. Here, we report on simulations of specific algorithms that allow to dynamically adapt the network to an optimal value of Tpoll during transmission. We place special emphasis on the role of the kind of traffic on the network (e.g., constant bit rate, burst, multimedia-like). In addition, we analyze the results of simulations by comparing these algorithms with the static allocation value defined in the Bluetooth specification. Our results provide insight on how to optimize the transmission of multimedia traffic over piconets at the lowest layers of the Bluetooth architecture.