Cells: A philosophical excursion (Part 1)
This is the first of a two-part excursion into the realm of philosophy as it relates to “cells.” Don’t worry, I’ll leave the important questions like “what is enlightenment?” to Emmanuel Kant and his successors. But since the evolution of LTE to LTE Advanced (LTE-A), it is worthwhile to ask: What is a cell?
In the beginning of cellular communications, there was the base station. Good network planning was constituted by deploying base stations in such a way that the coverage the base stations provided was ubiquitous, yet not redundant. Boy, have things changed! Welcome to 2012, a world of carrier aggregation (CA) and coordinated multi-points (CoMP): the world in which this blog takes place. Techniques such as CA and CoMP have dramatically changed the notion of a cell.
In early releases of LTE, bandwidth was limited to 20 MHz whereas in LTE-A, the theoretical bandwidth can now be as large as 100 MHz. However, each “chunk,” called a component carrier, cannot have more than the legacy 20 MHz bandwidth. Larger bandwidths are achieved by aggregating component carriers. From a protocol standpoint, though, a component carrier is essentially one cell.
Traditionally, we think of a cell as a coverage area. Users receive data from, and transmit data to, one base station. In CoMP, multiple geographically dispersed base stations coordinate the transmissions and in the extreme case, users can receive data from, and send data to, multiple base stations simultaneously. To further complicate matters, CoMP can be implemented completely transparent to the mobile user and a set of base stations can appear to a user as one cell.
What does this mean to network operators? They need scalability and an architecture which delivers the highest performance during peak hours, when multiple cells are aggregated through CA and CoMP, respectively. At the same time, when the network is operated in more traditional ways, where each base station provides one cell for ubiquitous coverage, hibernation modes are required for unused cores or even entire chips.
Today, I’ve taken the view of the network to ask: What is a cell? In part 2, I’ll ask the same question from an end-users’ perspective and look at a trend which began in LTE Release 10, but has really come to fruition in LTE Release 11. Does anyone out there want to venture a guess on what I’m referring to?
For more posts by TI multicore experts, visit www.ti.com/multicoremix_ecn