In response to operator's demands for performance improvement, MAS takes different forms in existing networks and in next-generation mobile broadband networks.

Operators rolling out services based on HSDPA and EV-DO technology are discovering the client data rate problem previously outlined. They are finding that their existing 3G network footprint creates “Swiss cheese” coverage for high data-rate services, and that their suppliers' client devices' performance varies widely from one model to the next.

Because there are significant constraints on deploying MAS-based infrastructure solutions in existing networks, 3.5G operators are urging their client device suppliers to develop MAS implementations on the client side. In response, Qualcomm has added receive diversity and some simple coherent-gain processing to its chipsets, and EV-DO data cards from Sierra (for example) have implemented two-antenna solutions on this basis. Others are working to catch up on diversity solutions, as well. To indicate what's coming next in this area, Figure 5 shows the output of detailed real-world network simulation (using a European operator's planning tool and network configuration) for ArrayComm solutions that add interference mitigation to the diversity+combining baseline established by Qualcomm. Even in lightly loaded networks, interference from the neighboring base station or a single dominant co-channel user can substantially degrade data rates at cell edges. Adding a second antenna to the client allows it to ignore a number of co-channel interferers, with significant net results — more than doubling data rates throughout the network in real-world conditions. These interference-cancellation techniques continue to advance in the 3GPP RAN4 discussions of future interoperability standards for 3G client devices.

Likewise, the 3GPP discussions of 3G long-term evolution (LTE) and the Mobile WiMAX profiles established in the WiMAX Forum process are both embracing the combination of orthogonal frequency-division multiplexing (OFDM) and MAS architectures.

Many approaches to MAS concepts have been attempted over the past 15 years. Some early trials involved large, expensive, and precisely calibrated arrays that in the end didn't work well, and some involved so-called “appliqué” solutions — aftermarket add-on boxes that yielded generally poor performance because of limited integration with the existing radio hardware and necessarily unsophisticated algorithms.

Previous 1 2 3 4 Next

RSS    Save to Del.icio.us  Digg This