= FlexCRM Project =

[[Image(CRM.png, 550px, align=right)]]

SDR presents a hard real-time computing challenge with varying (computing) system conditions. The computing resource management framework thus needs to track the states of the computing resources and allocate or reallocate them as needed.

An SDR processing chain, SDR application or waveform, is the part of an SDR transceiver that is implemented in software. It can be understood as a set of concurrent processes that continuously process and propagate real-time data. An automatic mapping process (computing resource manager) dynamically assigns software modules to hardware resources, while meeting all computing system constraints.

The computing resource management framework features the computing system modeling and computing resource management modules (Figure 1). Figure 2 illustrates ALOE's time management principle. The pipelined execution pattern facilitates the synchronized execution of the waveform modules on distributed computing resources. The SDR computing system modeling then captures the computing resources and the computing requirements on time slot basis. Million operations per time slot (MOPTS) and mega-bits per time slot (MBPTS) are used for modeling the processing and interprocessor data flow capacities and requirements.

The framework offers general-purpose mapping algorithms and a customizable cost function ([attachment:FlexCRM_July11.pdf FlexCRM_July11]). The cost function implements the computing resource management objective and guides the mapping process under the given computing constraints. The computing resource management framework--mapping API, algorithms and cost function, and simulation test suite--can be downloaded following [attachment:CRMframework_July11.zip this link], which contains the C source files. The [attachment:CRMtools_July11.zip CRM tools] permit executing the computing resource management framework from Matlab, following the mapping process, and analyzing the results. [wiki:ALOEedu ALOE Sessions 7 and 8] introduce the framework and its tools.

We currently investigate how to minimize the pipelining latency. We therefore examine new cost functions and evaluate their performance in different computing resource management scenarios. We, furthermore, address the scalability of the mapping algorithms with the objective of applying them to large-scale computing systems.


[.. Back to Main Page]