Power and Area Efficient VLSI Architectures for Communication Signal Processing

The ever increasing demand for data-centric wireless connectivity has inspired the belief of complex signal processing algorithms like those utilized in MIMO communication. MIMO stands for Multiple-In Multiple-Out, pertaining to the very fact that when a packet is transmitted into the channel it transmitted via multiple inputs and when it comes out of the channel it’s received on multiple antennas. Major techniques to attenuate power or area (as a dominant measure of cost) are architectural in nature. With technology scaling, designers have more options in selecting supply voltages and transistor thresholds additionally to varied circuit design techniques just like the utilization of sleep modes. The goal is to therefore develop a strategy that simultaneously minimizes power and area for given throughput and latency constraints. There are multiple ways and options in terms of the architectural methods to choose from depending on the kind of output the user would prefer.

· In parallelism and time multiplexing, parallelism combined with adjustment in supply voltage improves the energy by slowing
down the clock and distributing computation over several
parallel branches. Time-multiplexing does the exact opposite: it reduces the area by repeating computation on an equivalent hardware unit, but needs a better clock rate and better supply voltage to take care of the throughput. Alternatively, it can reduce performance by maintaining the clock rate.

· Loop retiming is a technique of distributing pipeline registers around recursive loops. A recursive loop is said to have occurred when a function, module or an entity keeps making calls to itself repeatedly, thus forming an almost never-ending loop. The goal is to assign the right amount of latency to basic functional building blocks and then distribute the pipeline registers inside the blocks such that all internal data path logic blocks lay at the same point.

What we saw here were some of the methods to be implemented for communication signal processing. But the design framework is just as vital for the same.

Matlab and Simulink based style surroundings permits theorists and implementers. Associate in Nursing algorithmic rule is entered solely once may be a graphical block type, that provides regular data-flow illustration and abstract read of design architecture. With technology-specific information for speed, power and space of purposeful blocks, algorithmic rule designers will explore the implementation house whereas remaining in Simulink environment to verify the algorithmic rule.