The wireless challenge

As competition continues to intensify for telecommunications service providers, the pressure to find more ways to decrease operational and maintenance costs while improving network performance is reaching an all-time high.

One way to reduce such costs is to decrease the amount of power needed to run networks. This has become increasingly critical in wireline networks such as fiber-to-the-curb and switched digital video architectures.

With the explosion of the wireless industry, decreasing power consumption and cost in these net- works is becoming equally crucial.

Traditionally, power consumption in wireless telecommunications has focused on the cellular telephone, for which talk time and standby time are key selection criteria. Now, however, wireless infrastructure equipment manufacturers are taking a hard look at how to add value to their next generation equipment by making it more power-efficient. This need has manufacturers looking to component suppliers to provide solutions that are significantly more power efficient than the components in current equipment.

The digital signal processor A key component used in both wireless terminals and infrastructure equipment is the digital signal processor (DSP), a specialized microprocessor that excels at performing multiply-and-add operations in the form of A = B * C + D. This type of operation is central to DSP algorithms that are used to manipulate and process voice, video and data signals. A rough measure of a DSP's performance is how many millions of multiply-and-add operations, or instructions, it can perform in one second.

DSPs are one of the fastest growing semiconductor market segments. They have been widely deployed in wireless infrastructure base station equipment, cellular phones and two-way paging network components. Housed within base station cabinets, DSPs are designed to support various wireless technology standards around the world, including code division multiple access, IS-54 and IS-136 in North America.

In a typical wireless network, the base station provides the link between the user and the rest of the network (Figure 1). Base stations are responsible for transmitting, receiving and processing all cellular transmissions within the coverage area, or cell. Cellular transmissions include voice and data traffic from mobile users, as well as timing, synchronization and control information that makes wireless networks function without interruption as users' signals travel within and between cells.

DSPs are used extensively in base stations to perform channel coding, equalization and encryption on the transmitted signals. The channel coding function involves packing transmissions with error correction information so that they are less susceptible to electromagnetic interference as well as interference from buildings, mountains and other structures.

Equalization is used in base stations to effectively eliminate the problems that can occur when a signal transmitted by a cell phone takes multiple paths to the base station antenna. This situation results in multiple versions of a transmission, each with a different signal strength, arriving at the base station at different times. If these signals are not processed properly, the result will be an unintelligible conversation.

Encryption is used in digital wireless systems to provide security during transmissions and to prevent eavesdropping of calls.

Base station controllers support 10 or more base stations and provide a link between base stations and the public network. Base station controllers are responsible for providing seamless handoff of calls from one base station to another as mobile users cross from one cell area to another.

In digital wireless networks, the controller also performs speech compression and decompression. Speech compression is used in digital wireless networks to provide more capacity than analog systems using the same amount of available spectrum.

The user's speech is compressed by the cell phone before transmission to the base station. The compressed speech is then received by the base station, which sends it to the controller for decompression before sending it onto the public network, which does not recognize compressed speech. In the return path, uncompressed speech from the public network is first compressed in the controller before being sent downstream to the appropriate base station and cell phone user. DSPs are widely used in base station controllers to perform speech compression and decompression.

Reducing power consumption Efforts to reduce power consumption in new generation infrastructure equipment will focus on the base station because of the high ratio of base stations to controllers-ten to one and higher. The signal processing functions performed in base stations require a significant amount of compute power, necessitating several DSPs to handle a single radio channel. Current base station designs use six or more DSPs to process one radio channel for a total of eighteen or more DSPs per typical base station.

A new generation of DSPs, including Lucent Technologies' DSP1620, is responding to equipment developers' requests for more power-efficient solutions. Introduced earlier this year, the DSP1620 can perform 120 million instructions per second-a three-fold increase in compute power over earlier versions.

This means equipment manufacturers will require as few as two DSPs to perform the same functions as six or more in today's equipment. And because each of these new DSPs consumes only as much power as each of the current DSPs, the net result will be a 300% reduction in DSP power consumption.

The new generation of DSPs requires only a 3-volt power supply, compared with the 5-volt supplies used in the previous generation. This allows manufacturers to use much smaller power supplies that consume significantly less power.

Another contributor to DSP subsystem power consumption is random access memory. In current designs, each DSP in the base station has a block of external memory to hold the code and data required to perform the signal processing functions. Another significant savings in power in the new generation of DSPs comes from incorporating the memory on the circuit board.

Other benefits Heat is a natural by-product of power consumption. Today's wireless systems use fans to generate airflow across the components and to exhaust the heat that builds within the equipment cabinets.

Because new generation DSPs draw less power, they contribute to a system that runs at a much cooler temperature. Macrocell base stations, which provide coverage over wide areas-up to 35 kilometers-will require less cooling, while smaller base stations such as microcells and picocells, which provide coverage for several kilometers to several hundred meters, will be able to eliminate the need for cooling altogether.

This will be especially significant in the years to come, as the deployment of small base stations will outpace macrocells. The smaller base stations will facilitate higher calling capacities within densely populated geographic areas, which will be increasingly important as penetration levels-especially in urban areas-increase.

DSP power savings will also have a trickle-down effect on battery backup systems designed to provide continued service in the event of a brownout or blackout. The result will be smaller, lower-cost backup systems.

Because of the ever-present demands on network service providers to reduce costs, wireless infrastructure equipment manufacturers will be forced to further reduce system power consumption beyond next generation designs. This will require component manufacturers to develop even more efficient solutions. The result will be a DSP powerful enough to handle the base station functions with just a single device. This DSP will operate at an even lower voltage and consume a fraction of the power of today's solutions.

Narciso Mera is Wireless Infrastructure Marketing Manager for Lucent Technologies' Microelectronics Group, Murray Hill, N.J.