DocumentCode
3474954
Title
TMS320C66x multicore DSP efficiency in radar imaging applications
Author
Vityazev, S. ; Kharin, A. ; Savostyanov, V. ; Vityazev, V.
Author_Institution
Fac. of Radio Eng. & Telecommun., Ryazan State Radio Eng. Univ. (RSREU), Ryazan, Russia
fYear
2015
fDate
14-18 June 2015
Firstpage
115
Lastpage
118
Abstract
Radar imaging is a very attractive application of digital signal processing widely used in modern space-, air- and ground-borne systems. Radar imaging algorithms are featured with very high computational requirements. New trends in processor platforms with a high level of computational power are very important for radar imaging implementation. TMS320C66x multicore DSP from Texas Instruments is one of a promising computational platform for radar imaging systems. It demonstrates good performance in embedded computing with low power consumptions. A test task implementation on TMS320C6678 multicore DSP is considered in this paper. The test task is described and implemented stage by stage on one core of TMS320C6678 device with an analysis of processing time limitations. 8-core implementation is considered then. It is shown that test task is done in 4.9 ms with 1 core and 0.99 ms with 8 cores implementation. The paper demonstrates an efficiency of TMS320C66x DSP platform in radar imaging applications and discusses the questions of an efficient multicore programming.
Keywords
multiprocessing systems; radar imaging; signal processing; 8-core implementation; TMS320C66x multicore DSP efficiency; Texas Instruments; digital signal processing; embedded computing; low power consumptions; multicore programming; radar imaging applications; radar imaging systems; Algorithm design and analysis; Arrays; Digital signal processing; Multicore processing; Radar imaging; Signal processing algorithms; Software; DSP; Radar Imaging; TMS320C6678; multicore; optimization;
fLanguage
English
Publisher
ieee
Conference_Titel
Embedded Computing (MECO), 2015 4th Mediterranean Conference on
Conference_Location
Budva
Print_ISBN
978-1-4799-8999-7
Type
conf
DOI
10.1109/MECO.2015.7181880
Filename
7181880
Link To Document