IEEE ComSoc CTN Special Issue on Towards The Open Basestation

CTN Issue: February 2015

By Alan Gatherer, CTN Editor in Chief  & Huawei Chief Technology Officer, Baseband on System Chip

Traditionally the Basestation in a cellular network covers approximately layer 1 through layer 3 of the network functionality and is seen as a closed box into which bits go in and waveforms come out, and vice versa. In logical network diagrams it is often portrayed as a tiny box at the edge of a large and complex network of wires, gateways, home locations, authenticators and so on. However, this belies the true cost of the Basestation equipment which can account for 40% or more of network capital expenditure [1]. The relative cost of Basestation equipment is increased by trends towards denser networks with smaller cell sizes, overlays, and more cooperation between cell sites using techniques such as Cooperative MultiPoint (CoMP), all of which reduce the utilization of the Basestation [2], because the traditional network is designed for worse case load at every point in the network. In reality the network is a much more dynamic, almost organic entity where patterns of equipment loading follow statistical rules. The term “tidal effect” [2] is used to describe long term user trends but short term user fluctuations also obey statistical rules and can be taken advantage of [3]. To compound this problem users patterns are also affected by trends in application use and changes due to events in the short term and street and restaurant popularity in the medium term. User movement also seems to obey statistical rules from biological systems [4] and this can be taken advantage of to come up with simple and intuitive results on how user movement affects Basestation resource use [5].

Above we describe the basestation as a collection of resources that is sampled and used in a statistical manner. We can take this further and state that the Basestation should be designed as a collection of services that are utilized in a statistical manner, with some probability of failure to orchestrate the required resources to achieve the desired results. This can be most easily implemented in a C-RAN architecture but is not limited to that kind of architecture. In [6] the argument is made that a “programmable infrastructure”, based on the principles of software Defined Networks (SDN), is the only answer to the diverse set of applications that future wireless networks will need to support. The emerging requirements for 5G cellular ([7] and many other places) seem to bare this ultra diverse requirement space out. In [9] they give an example of a “user centric” network in which a collection of WLAN access points coordinate to create virtual access points that follow users around. This kind of flexible and user centric networking is likely to become important in 5G and will require a very flexible Basestation. As we are now orchestrating and service chaining a mixed criticality, real time system, it becomes important to understand the real time guarantees of service [8][3]. This is still a new area of research for wireless cellular that sits at the intersection of scheduling, software defined networking, and wireless modem design.


[1] “Relation Between Basestation Characteristics and cost Structure in Cellular Systems”, Johansson, K. et al, 15th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2004. 5-8 Sept. 2004. Page(s):2627 - 2631 Vol.4

[2] “C-RAN, The Road Towards green RAN”, White Paper Version 2.5, Oct 2011, China Mobile Research Institute.

[3] “Fog-RAN: Hardware Resource Sharing in Aggregated Baseband Processing Systems”, Zhu et al, Design Automation Conference, 2014.

[4] Injong Rhee; Minsu Shin; Seongik Hong; Kyunghan Lee; Seong Joon Kim; Song Chong "On the Levy-Walk Nature of Human Mobility",  Networking, IEEE/ACM Transactions on, On page(s): 630 - 643 Volume: 19, Issue: 3, June 2011.

[5] “Towards Understanding the Fundamentals of Mobility in Cellular Networks”, Lin et al, Globecom 2012 and

[6] “OpenRadio: a programmable wireless dataplane”, Proceedings of the first workshop on Hot topics in software defined networks, August 13 2012. And

[7] “4GAmericas’ Recommendations on 5G Requirements and Solutions”,

[8] “CloudIQ: A Framework for Processing Basestations in a Data Center”, Bhaunik et al, Mobicon 2012.

[9] “BeHop : A testbed for dense WiFi networks”, Yiakoumis et al, Wintech 2014, Maui, USA, 2014.

Leave a comment

Statements and opinions given in a work published by the IEEE or the IEEE Communications Society are the expressions of the author(s). Responsibility for the content of published articles rests upon the authors(s), not IEEE nor the IEEE Communications Society.