Font Size: a A A

Study And Implementation Of A CDMA Baseband In WSN

Posted on:2008-01-20Degree:DoctorType:Dissertation
Country:ChinaCandidate:M S WuFull Text:PDF
GTID:1118360242958314Subject:Microelectronics and Solid State Electronics
Abstract/Summary:PDF Full Text Request
The Wireless Sensor Networks (WSN) consisting of densely distributed nodes is a integrated autonomous information system which combines sensing ,signal processing,decision capability.Sensors are likely to lie on near the ground as opposed to a person holding a cell phone,so near-ground wave radio-wave propagation is vital to WSN . Batteries power typically hosts in WSN, so low power design is very critical for WSN node. Furthermore,sensor nodes are limited in power,computational capacities,communication capacities and memory. The dissertation reached for the solution aimed at the problems upwards. Because of the constrained energy and size , the system on a chip(SOC) design including baseband processor and microprocessor for the protocol stack of WSN is a direction of the development.For decreasing the energy consumption and size of the nodes,the paper involved in the CDMA baseband processing and circuit implementation to restraining multipath fading with burst mode transmission.In this dissertation,the author simulated the CDMA baseband processing algorithm with Cocentric System Studio EDA platform of Synopsys company.A dedicated pilot channel,which is separated from data channels,is used for the purpose of channel estimation.Pilot channel estimation error due to multipath access and multipath interference and AWGN in multipath fading channel is studied. System and maximal ratio combining performance is evaluated by means of bit error rate(BER).The analysis showed that the error of channel estimation significantly degrade the system performance and can be effectively suppressed by low pass filters.The power ratio of pilot to data channels is studied and the optimal value is given.Considering the WSN nodes are used widely in various wireless channel scenarios,a new adjustable finger search window length RAKE receiver scheme is developed. The paper investigated optimized design of the circuit clock tree to reduce clock skew which affect the circuit function. To save the chip area and decrease the circuit power, Considering the coefficient symmetry of the filter and the oversample trait , a novel design and implement of baseband pluse shaping filter based on Look-Up Table is proposed which can reduces the chip resoureces much compared to the conventional design way.In the Viterbi decode scheme,some efficient methods such as the in-place modification operation for pathmetrics,the saving of path transition and the circle access for path surving are presented. By using these methods, the RAM size needed for saving metrics and paths and the power consumption are decreasedThe theoretic analysis and deduction in this dissertation help to understand the essence of channel estimation error and enhance channel estimation in WSN. The given power ratio of pilot to data channels make a guidance to the practical project. Furthmore,the implementation of the baseband processor applied in the design of WSN nodes laid a foundation for the succeeding SOC R&D work.
Keywords/Search Tags:Wireless Sensor Networks, Multipath Interference, SOC, Low Power Design, Channel Estimation, RAKE Receive
PDF Full Text Request
Related items