Rakon have dominated the market for femtocell oscillators since their inception with well over 85% market share. Here we look back at how they established their position, and look ahead to the changing requirements for LTE.
The crystal oscillator in a basestation is typically an expensive component. An OCXO (Oven Compensated Crystal Oscillator) can cost several hundred dollars alone, far exceeding the total target cost of a small cell. During the early stages of femtocell development, Rakon pioneered the use of TCVCXO (Temperature Controlled Voltage Controlled Crystal Oscillators) which met the price point while keeping the tight tolerances required.
The company was founded in 1967 in New Zealand by Warren Robinson, and is now run by his son. Their scope ranges from mass market GPS receivers to a wide range of electronic equipment with timing needs. Their foresight of a large market combined with a reputation for high quality, and investment in high production capacity, has allowed them to grow. Shipments are in the millions of parts per month.
They acquired C-MAC, a European company in 2007, which had an in-house ASIC design team. This team developed the Pluto chip, with combined ASIC and crystal components in a single package that met Stratum 3 level performance and quickly became adopted as the de-facto industry standard.
The company has become globally distributed, with several design centres and multiple manufacturing sites in China, New Zealand, India, UK and France. The femtocell business is run from their Lincoln, UK office but the product can be manufactured can be from one of several sites – it's genuinely a dual-site operation. This allows the company to rapidly respond to market demand, ramping up production to match orders.
3G Femtocell Oscillator
Rakon's TCVCXO met the 100 parts per billion (ppb) tolerance at a price point that an OCXO couldn't match. Their in-house ASIC design is the major differentiator, but they've also established a strong reputation for quality. Ageing can be an issue for quartz crystals, and although the nominal voltage can be adjusted to compensate at any one temperature, Rakon also ensure that the stability is maintained over the life of the product and the range of voltages.
Their product is integral to most 3G femtocells today, although other vendors are entering the market and can be the choice of second/alternate supplier.
LTE femtocell requirements are more demanding
This new 4G technology has different requirements from 3G. It's not just the accurate frequency that must be maintained, but there is a phase noise mask to meet. This affects the achievable maximum throughput at short range.
LTE has two distinct modes:
- FDD which operates on paired spectrum, sending and receiving on different frequencies
- TDD where all cells alternately and simultaneously send and receive on the same spectrum
For FDD mode, GSM/UMTS and LTE can all run asynchronously, with clocks running at the same frequency but not phase synchronised. LTE does have an option for simulcast, known as MBNS (Multimedia Broadcast Multi-mode Service), which if used would require phase synchronisation of a few microseconds tolerance. Tightly integrated HetNets, where uplink and downlink signals are from different cells, also require phase synchronisation.
For TDD mode, there is a hard requirement for 3 microsecond phase sync.
Phase sync to this accuracy can't simply be achieved by standalone femtocells and DSL alone, and would need one of GPS, synchronous Ethernet or external macrocell network listen. If a fixed DSL network provider really wanted to help, it would be technically possible to design a solution that was integrated with the DSL modem and DSLAM to deliver synchronisation using the DSL physical layer in a broadly similar way to Synchronous Ethernet.
Rakon have seen demands for a new market segment that fits between the traditional macrocell OCXO and the low cost 3G femtocell TCVCXO. Their new Mercury product is an ASIC driven OCXO with heating and thermal management all integrated onto the silicon. Typically an OCXO is required for high stability, low phase noise and long holdover periods. Issues with conventional products include high power consumption, size and cost. There is also a need for additional discrete components .
The company report a lot of interest from general telecoms customers, especially those looking for outdoor/public access metrocells. The Mercury product is suitable for both 3G WCDMA and LTE, due to its excellent phase noise performance. It will fill the gap (in optimum oscillator technologies) between commodity residential femtocells and large macrocells, ideally suitable for the growing range of metrocells
A burden of responsibility
Rakon see themselves as having a huge responsibility to the femtocell and small cell industry, and believe it very important to be able to honour commitments to deliver on time. With good stock levels and short lead times, their ability to manufacture in multiple locations means they can ramp up quickly to meet rapid growth in demand.
Long term membership of the Small Cell Forum (they were one of the earliest members of the Femto Forum) has provided good insight and awareness of what operators are doing. This has helped validate the order pipeline and avoid double counting. The company has also been able to contribute to the White Paper on femtocell network synchronisation available from the Small Cell Forum website.
More product information can be found on Rakon's Website: