In a market where standards and frequency bands are constantly being introduced, original equipment manufacturers (OEMs) are having to introduce new products ever more quickly, simply to keep up.
The average femtocell maker has to contend with 3G, Wimax and long term evolution of 3G (LTE), all in different frequency bands for different locations, such as the availability of the 700MHz band in the US, for which a new transceiver, power amplifier and antenna are needed each time.
Recently, rather than starting from scratch each time, OEMs are beginning to look to programmable solutions to give them a head start. The idea is to make a small cell basestation that can be programmed to work on a different standard or at a different frequency, depending on where/how the end product will be deployed.
MEMS switches as a solution
Baseband ICs are increasingly becoming programmable, but making a multi-band, multi-standard front end has been more of a challenge. Until quite recently it was thought the best way around this would be to use a bank of RF micro electromechanical system (MEMS) switches, switching between several different RF front ends, as no-one thought the RF silicon guys could do it.
However, opinions are changing as companies develop programmable multi-band, multi-standard transceiver ICs, which can be configured digitally to operate as Wimax, 3G or LTE transceivers at frequencies between 375MHz and 4GHz, with user-selectable bandwidths.
As a result the programmable silicon approach can offer real benefits to OEMs and system designers. A programmable transceiver, provided it is sufficiently frequency-agile, removes the need for individual transceiver chips for each of the different standards or geographical territories. The chips can be reprogrammed rapidly and simply to function as a transceiver for various network configurations, bandwidths and data rates.
This means OEMs do not have to go back to the silicon vendors each time a new spectrum allocation appears and ask for the design and implementation of yet another transceiver IC, which would inevitably impact upon time to market.
Once a product has been developed around a particular transceiver, this maximises design reuse, again shortening the production cycle. These factors allow OEMs to quickly develop new products to meet the market need.
Programmable siliconThis is all well and good, but programmable silicon will never grab the attention of femtocell makers unless it is also competitive on cost. The words ‘low-cost’ and ‘programmable silicon’ are not usually seen in the same sentence, but if you consider the problem in broader terms than just the per-unit cost, a convincing proposition does begin to emerge.
Firstly, OEMs will be able to take advantage of economies of scale, as one transceiver or baseband chip will be used in multiple product lines.
Secondly, there are also significant implications for OEMs’ inventory. It will only be necessary to purchase one type of transceiver and one type of baseband chip for an entire product range of small cell basestations.
This is particularly useful for global OEMs shipping to different geographical locations. So if demand goes up in one country, you already have the silicon you need, no matter which country it is.
These developments in programmable baseband and transceiver ICs are driving the small cell basestation sector towards a single bill of materials. This is the goal: a single end-product whose hardware can be fully reprogrammed to support any standard or frequency.
We are not there yet, as power amplifiers and antennas still have a way to go, but the advantages in terms of logistical savings and time to market are apparent. It can only be a matter of time.
Ebrahim Bushehri is CEO of Lime Microsystems