Consumer femtocells and their higher power cousins, enterprise and public access femtocells, provide coverage in hard-to-reach areas. But they do not address the mobile data capacity explosion. Why? Because they cannot be used in places where the demand for mobile data is actually exploding!
Spidercloud’s Amit Jain is speaking today at the LTE LATAM 2013 conference, taking place at the Windsor Barra Hotel, Rio de Janeiro, Brazil.
The demand for mobile data is highest in places where hundreds or thousands of people congregate, such as large shopping centres and large office buildings. Using a single small cell, irrespective of its power or capacity, will not help operators meet the demand for data. All that the operator will get is dissatisfied subscribers, who can see five bars of coverage, but merely get a few hundred kilobits of data.
To address the mobile data explosion, operators need a small cell system that enables them to:
- Build a dense small cell network inside buildings, with numerous small cells
- Easily add more small cells as more smart phones and more apps come on the network
- Provide consistently high throughout, and consistently low call drop rates
- Deploy this small cell network in hours or days, with technicians who are not cellular gurus
This is a tall order. The indoor RF environment, especially in large multi-storey buildings is very challenging. In a dense deployment, a handset can see several small cells at the same time. Because of fast fading, a handset may handover from one cell to another several times a minute without moving at all.
So, is a dense small deployment not possible? Yes and no. It depends on the architecture adopted. Broadly, four architectures have been proposed in the industry:
1) Femtocells connected to a Home Node B Gateway (HNB-GW) with hard handover
2) Small cells connected to a Home Node B Gateway (HNB-GW) with soft handover using “Iurh”
3) Pico-cells connected to a traditional 3G Radio Network Controller (RNC)
4) Small cells connected to a small local controller. Local controller connects to the core network as single HNB.
The first option, hard handover of femtocells, has been trialled by many operators and most agree that it is not practical to deploy more than 5-10 femtocells in a large building.
Many suppliers who initially proposed the first architecture are now moving to the second architecture. They are implementing soft handover using a variation of the Inter-RNC handover protocol called ‘Iurh’. Since soft handover requires synchronization between small cells, some suppliers are building small cells with expensive oven-controller oscillators. All handover signaling goes over the backhaul link and can become a significant expense. And there is no way for an operator to locally offload data traffic without breaking inter-small cell mobility. Products based on this architecture are currently in development.
The third option is using pico-cells connected to a RNC is another way to do soft handover between small cells. This architecture is often offered by macro cellular infrastructure suppliers, who are able to scale down their macro NodeBs and reuse existing RNCs. It can be attractive if an operator requires a small number of small cells, but in the case of high density deployments, the cost of RNC ports can add up. Further, this architecture does place very stringent requirements on backhaul, and it unclear how SON functionality will be implemented.
In the fourth architecture, all small cells in a building connect to a small local controller over Ethernet. This controller is responsible for managing mobility, interference and SON. It aggregates all the traffic and connects to a HNB gateway as a single HNB would using standard Iuh signaling. All inter-small cell mobility events stay inside the building, and do not load the backhaul link or the HNB-gateway. The local controller acts as the master-clock and synchronizes all the small cells, eliminating the need for expensive oscillators in every small cell. If an operator wants to offload data traffic locally or integrate with enterprise applications, it can do so using the local controller. Some innovative operators are working on innovative enterprise applications that use the network intelligence that can be accessed at the local controller.
SpiderCloud’s 3G small cell solution is based on the fourth architecture. Operators have used it to deploy as many as 65 small cells in a 16-storey office building, with thousand of subscribers and hundreds of thousands of inter-small cell handovers daily and the technology is now ready to provide coverage, capacity and new applications in even larger buildings.
Amit Jain joined SpiderCloud in September 2011. Prior to SpiderCloud, Mr. Jain was vice president of marketing, sales and service for Airvana’s CDMA femtocell business and his tenure at Airvana spanned ten years, including the company’s inception. At Airvana, Mr. Jain held several leadership roles in marketing, business development and sales for 3G EVDO macro cellular products and femtocells. Prior to Airvana, Mr. Jain held both technical and business positions at Qualcomm, Ericsson, and McKinsey & Company. He holds an MBA from MIT’s Sloan School of Management, an MS in Electrical Engineering from University of California at Irvine, and a B.Tech in Electrical Engineering from the Indian Institute of Technology, Bombay.