Techniques to Improve Blueooth Performance in Interference Environments

Techniques to Improve Blueooth Performance in

Interference Environments

Nada Golmie and Nicolas Chevrollier

National Institute of Standards and Technology

Gaithersburg, Maryland 20899

Email: nada.golmie@nist.gov

Abstract--Bluetooth is a radio technology for Wireless Personal Area Networks

operating in the 2.4 GHz ISM band. Since both Bluetooth and IEEE

802.11 devices use the same frequency band and may likely come together in

a laptop or may be close together at a desktop, interference may lead to significant

performance degradation. The main goal of this paper is to propose

solutions to the interference problem consisting of power control adjustments

and scheduling policies to be implemented by the Bluetooth device. Simulation

results are given for selected scenarios and configurations of interest.

Keywords--Bluetooth, Interference, Power Control, MAC scheduling

I. INTRODUCTION

The Bluetooth [1] technology is an emerging short range cable

replacement protocol operating in the 2.4 GHz ISM band.

Since both the Bluetooth and the IEEE 802.11 [2] protocols

operate in the 2.4 GHz, it is anticipated that interference may

severely degrade the performance of both systems.

Our goal is to propose solutions to the interference problem

pertaining to the Bluetooth radio operating in proximity to an

IEEE 802.11 network. We assume that the source of interference

to the Bluetooth system is an IEEE 802.11 system operating

in a direct sequence spread spectrum (DSSS) mode. In the

rest of this sequel, the terms IEEE 802.11 DSSS and WLAN

will be used interchangeably.

We investigate two techniques aimed at alleviating the interference

problem for Bluetooth. One technique is based on controlling

the transmitted power and keeping it proportional to the

signal-to-interference ratio (SIR) measured at the receiver. The

other technique takes advantage of the frequency hopping sequence

of Bluetooth and uses scheduling with the aim of avoiding

interference. Simulation results for scenarios of interest are

discussed. Performance is measured in terms of the mean access

delay, the probability of packet loss, and the transmitted

power.

This paper is organized as follows. In sections II and III,

we describe the distributed power control algorithm and the

scheduling mechanism respectively and give numerical results.

Concluding remarks are offered in section IV.

II. POWER CONTROL

Given that some devices provide the ability to dynamically

modify their transmission power, we would like to investigate

the dynamics of a power control (PC) strategy as a means of

alleviating the impact of interference.

We use a distributed algorithm to implement a PC procedure.

The basic idea is to adjust the interference level in the system to

no more than what is needed. We assume that the receiver does

not have any knowledge of other systems except for the system

it is communicating with. Interference from other systems is

measured in terms of the SIR level at the receiver. Note that SIR

is a wide-spread link quality measure and has been used in many

previous studies for power control and dynamic channel allocation

for interference limited systems [3] [4] [5]. The power

update algorithm works as follows. Initially, P0 = Pmax, then

every update interval U, the power at the transmitter, P(t + 1)

is updated as follows:

P(t + 1) = min(Pmax; max(Pmin;

t

SIR(t)

 P(t)) (1)

where  (t) is the target SIR and SIR(t) is based on an average

value over many measurements. The power update rule

takes into consideration the SIR(t) statistic measured at the receiver

side. The receiver can then relay this information to the

transmitter every update interval U.

Implementation Considerations Although the exact details

of a power control algorithm have been left undefined for the

most part, the Bluetooth specifications have included the necessary

hooks in the protocol in order to implement a power control

algorithm. Furthermore, the Bluetooth specifications classifies

devices into three power classes as summarized in Table I

TABLE I

BLUETOOTH DEVICE POWER CLASSES

Power Class Maximum Output Power Minimum Output Power

1