Font Size: a A A

Optimized Multiprocessor Implementation Of LTE-Advanced User-Plane Protocol

Posted on:2014-06-14Degree:MasterType:Thesis
Country:ChinaCandidate:J ZhangFull Text:PDF
GTID:2268330422463263Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
As one of the fourth-generation(4G) mobile communication technology dominated byChina, TD-LTE-A has been formally identified as the fourth-generation mobilecommunication international standard by ITU. By taking advantage of technologies suchas multi-antennas enhancement, coordinated multi-points transmission, and relaytechnology, TD-LTE-A system greatly enhances the spectrum utilization efficiency of theair interface as well as the system peak throughput. Due to the processing delaycharacteristics and system throughput requirement, the implemention ofhigh-performanced comprehensive test instrument will be up against great difficulties.Through the modern instruction architecture and advanced cache management technology,the X86SMP platform can meet the demand for high-performanced implemention ofTD-LTE-A protocol stack.Firstly, in order to take full advantage of parallel processing benefits of X86SMPplatform, several problems such as dynamic load balancing, inter-processors data sharingand cache consistency should be resolved efficiently. Considering the features of X86SMP architecture, the difficulties in implemention of TD-LTE-A layer-2protocol designcan be summarized as three aspects of multi-cores protocol stack model, efficientinter-core communication, as well as high-performanced memory management.Secondly, in the interest of achieving much better packet processing performance ofTD-LTE-A user plane, we approached three aspects of performance improvementtechnical schemes including interrupt optimization, compressed inter-processor datacommuting, and incremental packet encapsulation. Based on the intensively analysis ofmerits and demerits within above-mentioned optimized methods, we provided the basicstructure of the framework.Finally, we put forward a high-performanced multi-processors TD-LTE-A protocolarchitecture based on the X86MESIF infrastructure. This architecture has the advantageof more reliable performance, easier validation, lower cost and plenary transplantflexibility. The simulation testing result shows that the optimized architecture caneffectively improve the system performance and fulfill the throughtput and delaycharacteristic requirement of TD-LTE-A comprehensive test instrument.
Keywords/Search Tags:TD-LTE-A, Multi-processors, Throughput, Delay characteristic, Zero-copy
PDF Full Text Request
Related items