Femtocell and picocell technology is a way of delivering fixed-mobile convergence as an easier alternative to dual-mode cellular/WiFi systems in that it can work with standard cellphones. It involves putting smaller base stations in offices (picocell) and homes (femtocell) to deliver indoor cellular coverage and backhaul the traffic over an existing fixed connection, which can be a corporate network in the case of picocell or a home broadband link for femtocell.
The industry is talking about the need for a sub-$100 femtocell for this technology to really take off, so component providers like Bath-based picoChip are seeking to drive out cost in order to enable a lower bill of materials.
Vicki Griffiths, product manager for femtocell development at picoChip, said that even with its current generation of silicon the company enables a $100 bill of materials, provided the orders are in volume but said its next generation will take the process further.
We already enable a lower BOM than our competitors because we do all the baseband and networking on a single chip, she said, whereas they use multi-chip architectures with one or two DSPs for the interface and radio network control and an FPGA for hardware acceleration.
The picoChip portfolio consists of three products: the 205 chip for both femtocell and picocell for cellular and WiMAX networks, which ships with an embedded ARM processor where its OEM customers will run an RNC stack; the 202, which is the dedicated femtocell chip; and the 203, which is only for picocell base stations and ships without an ARM processor. The 203 is designed for scaling up. It uses an external host processor and several can be deployed using a single host processor for the RNC stack.
The developer’s competitors offering multi-chip products include DSP vendors such as TI, Analog Devices, and STMicro, all of which use FPGAs from companies such as Altera and Xilinx. Griffiths said picoChip has between 70% and 80% of the market for femtocell silicon.
