Dynamic baseband processor allocation in multi-radio base stations : A preliminary study

Abstract

Acting on the declining profitability trend of mobile data, operators have been trying to reduce network operational expenditure. Currently, different base stations are deployed for different radio access technologies (RATs). The advancement of technology has made base station components, like radio transceiver systems and baseband processors, reconfigurable to work with any RAT paving the way for multi-radio base stations that can handle multiple technologies with the same equipments.

This thesis looks at the possibility of having a pool of baseband processors that can be dynamically configured to the incoming traffic. This would be optimal as more processing power can be allocated to the traffic that needs it. This thesis is a preliminary study of the possible performance improvement that can be achieved by dynamically reconfiguring and allocating baseband processors to the incoming traffic.

A simulation model is developed using suitable traffic generation methods, understanding the air interface capacity and using a queueing theory approach to model the two processor allocation scenarios. LTE and WCDMA traffic are considered.

As a result, it is observed that the dynamic allocation scenario shows better performance than a static allocation of processors.