Power saving approaches for use with bluetooth: modelling and analysis of new protocols

The variety of Bluetooth applications has increased greatly in recent times. The energy consumption in such applications has emerged as an important problem. The research presented in this dissertation investigates the low power operations of current Bluetooth Basic Rate/Enhance Data Rate (BR/EDR) technology and explores several novel power saving optimization approaches for applications. In order to manage power consumption in these applications, certain features in Bluetooth BR/EDR are provided to allow low-power operations, e.g. using various operation modes and packet handling processes. As such, the research focuses on Bluetooth BR/EDR technology and improving power consumption by a packet transmission e ciency protocol and optimizing design of new operation modes. Firstly, a Packet Reassembly and Segmentation Protocol (P-RASP) in the Bluetooth baseband is proposed to operate during the idle/sleep interval duration in Bluetooth controllers. The protocol will re-assemble small host controller interface (HCI) data packets in the transmit bu ers to a larger one, so that the BT link manager can assemble a larger baseband packet type with full payload. Secondly, the research proposes a new strategy for reducing power consumption by improving the polling operation. The new approach uses a set of three di erent polling intervals in the Bluetooth BR/EDR controllers, whereby the controllers can choose the intervals and link state transfers from active to idle adaptively based on a common algorithm. The simulation results show this approach has very low average end-to-end packet delay and is easier and more exible in setting the parameters than Sni mode. Given the common algorithm or state-transition rules, a system model was established based on the Hidden Markov Model (HMM). The analysis shows the HMM can be a common model to analyze state-transition issues and be used to design and develop more e cient low power modes for Bluetooth in the future. The corresponding HMM utilization can be applied for the established system model. Finally, the design employs a M/G(M/M)/1/N queueing model and proposes a cross-layer approach to transmit data with the Hold mode in a low rate Bluetooth- based Medical Body Area Network. By using the proposed approaches and protocols above, the power consumption in Bluetooth can be signi cantly reduced. The results of this research may aid research and development teams to model, analyze and design a new operation mode to optimize power saving and improve e ciency for special Bluetooth applications.