What does picoChip do?

picoChip is a fabless semiconductor design company specialising in baseband processing chipsets for mobile network infrastructure. It has pioneered single chipset solutions for femtocells, with reference designs for HSPA, TDS-CDMA, LTE and WiMAX. Today its revenues arise mainly from 802.16e Mobile WiMAX equipment, where its chips are used in many vendors “rack and blade” WiMAX cellsites.

picoChip’s core expertise is in the baseband modem signal processing – clearing up radio signals received after being messed up through the transmission medium. This needs a lot of MIPS of processing power.

Their proprietary multicore DSP has enormous scalability, mixing hundreds of DSP cores and ASIC cores into a single System-on-a-Chip to meet specific market requirements.

Small cell radius is the only way to dramatically increase mobile network capacity

“Femtos are about a business model of small cell radius” he says. “In-building penetration has always been a problem for mobile network operators.  Hence the move to radio schemes which have high tolerance to loss versus high performance, speed and capacity”.

You may have heard of Moore’s law, where the processing capacity of microprocessors doubles approximately every two years. Or even Parkinson’s law which states that work expands to fill the time available and can also be applied to spectrum - however much capacity you have, usage will grow to fill it (which seems to be happening with 3G mobile broadband networks today!). Pulley refers to Cooper’s Law (Cooper designed the first mobile phone systems) which refers to the number of voice or data conversations which can be supported by all useful radio spectrum within the same area. This has doubled every two and half years – for over 100 years.

Doug’s point is that the capacity of mobile networks is increased much more dramatically through smaller cell radius rather than any other techniques. Smaller cells improve RF conditions – each supports fewer users with much lower interference levels.

[Ed Note: We’ve been saying this too. See The thirst for wireless data and Offload to femtocells ]

Femtocells could revolutionise outdoor public networks

“The Femtocell concept started out as a domestic use access point. We believe they could have a revolutionary impact on how 4G mobile networks are deployed because they can cover large areas much more cheaply than using traditional macrocells.”

[Ed Note: You’d still provide wide area coverage using existing 3G macrocells, 4G would be high capacity infill only in those areas requiring it]

Do operators need outdoor wide area 4G coverage?

“LTE is likely to be used as extra data capacity in highly congested areas, mainly indoors. Users will have USB dongles capable of both 3G and 4G. LTE femtocells would be found where data traffic is very high and devices would fall back to 3G when outside. This approach is much cheaper compared to installing smaller numbers of expensive LTE macrocells outside.

Since the benefit is primarily for the operator, they would have to manipulate user behaviour and steer customer adoption by pricing combined 3G/4G USB dongles appropriately. Operators are likely to take a firmer stance on LTE than 3G, making sure it’s in the operator’s interests.”

Metro-Femto rollout

The economics of a large scale outdoor femtocell rollout are quite different from the larger outdoor macrocells.  Kenny Graham, head of new technologies and innovation at Vodafone, is a keen proponent of what he terms “MetroZones ”. Hundreds of femtocells could be deployed in the same area previously covered by a single cellsite. Tradeoffs will need to include:

• Backhaul costs. Each LTE femtocell will need its own broadband backhaul link. This is likely to be wired, rather than wireless, and will want to be able to deliver the full data rates achievable – suggesting either fibre, Ethernet or ADSL2+.
• Power. Each LTE femtocell will need a mains power supply, either provided by site owner at cost.
• Site rental costs. Will the site rental for multiple femtocells (which would be deployed across properties owned by several companies be cheaper/more expensive than a single site?). This is simplified if a common owner is involved - such as lamp posts.

Could HSPA Femto’s be used outdoors in the same way?

 “They could do”, Doug says, “ but LTE has extra spectrum. HSPA is already in place in the macro network and provides the umbrella wide area coverage. LTE would complement that in high usage public areas.”

Don’t WiFi hotspots already provide this alternative high capacity infill? Look at ATT’s recent purchase of Waypoint, giving their data users access to 20,000 WiFi hotspots nationwide.

“The percentage of the metro areas covered by WiFi is still very low. How many times have you searched around for a WiFi hotspot and not found one easily?”

Who will deploy LTE first?

“The CDMA2000 operators are likely to be early adopters of LTE. They have been very involved in developing the standards – CDMA to/from LTE handover mechanisms are already standardised.

Although picoChip doesn’t provide productised CDMA reference designs for our chipsets, at least one femto vendor has developed their own using our platform. This could also be done for a dual-mode CDMA/LTE product.

We see a strong desire in operators for metro-femtos, and many operators have asked their usual suppliers to provide them.”

Comments (0)