Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1329696
C. Merkwirth, Jochen Bröcker, M. Ogorzałek, J. Wichard
In this paper we propose to use the discrete-time cellular neural network (DT-CNN) in a finite iterate mode. In such a mode of operation no special requirements on template stability properties are needed. We propose a constructive back propagation based algorithm for template design. For a given number of iterations we can find optimal sequence of templates for a given problem to be solved. Our novel approach is demonstrated by a design of a digit recognition DT-CNN.
{"title":"Finite iteration DT-CNN - new design and operating principles","authors":"C. Merkwirth, Jochen Bröcker, M. Ogorzałek, J. Wichard","doi":"10.1109/ISCAS.2004.1329696","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1329696","url":null,"abstract":"In this paper we propose to use the discrete-time cellular neural network (DT-CNN) in a finite iterate mode. In such a mode of operation no special requirements on template stability properties are needed. We propose a constructive back propagation based algorithm for template design. For a given number of iterations we can find optimal sequence of templates for a given problem to be solved. Our novel approach is demonstrated by a design of a digit recognition DT-CNN.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"1 1","pages":"V-V"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82976888","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1329480
B. Falkowski, C. C. Lozano, S. Rahardja
In this paper, an algorithm to generate disjoint cubes representation of a multiple-valued logic function is presented. The algorithm converts an array of non-disjoint multiple-valued cubes to an array disjoint multiple-valued cubes by applying different multiple-valued cube calculus operations. Efforts to reduce the calculation time and the number of resulting disjoint cubes have been integrated into the presented algorithm. Experimental results of the algorithm for several test files are also shown.
{"title":"Generation of disjoint cubes for multiple-valued functions","authors":"B. Falkowski, C. C. Lozano, S. Rahardja","doi":"10.1109/ISCAS.2004.1329480","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1329480","url":null,"abstract":"In this paper, an algorithm to generate disjoint cubes representation of a multiple-valued logic function is presented. The algorithm converts an array of non-disjoint multiple-valued cubes to an array disjoint multiple-valued cubes by applying different multiple-valued cube calculus operations. Efforts to reduce the calculation time and the number of resulting disjoint cubes have been integrated into the presented algorithm. Experimental results of the algorithm for several test files are also shown.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"1 1","pages":"V-V"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85413784","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1328715
M. Furtado, P. Diniz, S. L. Netto, T. Saramäki
Orthogonal frequency division multiplexing (OFDM) is a transmultiplex (TMUX) technique characterized by a low computational complexity and reduced inter-symbol/inter-carrier interference (ISI/ICI). Recently, new TMUX structures based on multicarrier modulation have been proposed aiming at even better performances. Such new structures include a cosine-modulated filter bank (CMFB) and a modified DFT filter bank (MDFTB). This work introduces design techniques for the CMFB and MDFTB structures with a large number of subbands (several hundreds or more), zero ISI, and small levels of ICI for most practical applications. The design is based on the concept of an Mth-band Nyquist filter, which inherently yields zero ISI and a highly selective filter bank, and in the FRM structure enabling one to synthesize FIR filters with a significantly reduced number of coefficients. The prototype example filter is optimized to minimize the resulting ICI. A design example is included to illustrate the benefits of the proposed technique in filter bank applications demanding a very large number of channels.
{"title":"Time-domain constraints for the design of FRM-based cosine-modulated and modified DFT filter banks with a large number of bands and zero intersymbol interference","authors":"M. Furtado, P. Diniz, S. L. Netto, T. Saramäki","doi":"10.1109/ISCAS.2004.1328715","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1328715","url":null,"abstract":"Orthogonal frequency division multiplexing (OFDM) is a transmultiplex (TMUX) technique characterized by a low computational complexity and reduced inter-symbol/inter-carrier interference (ISI/ICI). Recently, new TMUX structures based on multicarrier modulation have been proposed aiming at even better performances. Such new structures include a cosine-modulated filter bank (CMFB) and a modified DFT filter bank (MDFTB). This work introduces design techniques for the CMFB and MDFTB structures with a large number of subbands (several hundreds or more), zero ISI, and small levels of ICI for most practical applications. The design is based on the concept of an Mth-band Nyquist filter, which inherently yields zero ISI and a highly selective filter bank, and in the FRM structure enabling one to synthesize FIR filters with a significantly reduced number of coefficients. The prototype example filter is optimized to minimize the resulting ICI. A design example is included to illustrate the benefits of the proposed technique in filter bank applications demanding a very large number of channels.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"71 1","pages":"III-189"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85904293","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1329465
Hailong Yao, Qiang Zhou, Xianlong Hong, Yici Cai
Crosstalk noise is one of the emerging issues in deep submicrometer technology which causes many undesired effects on the circuit performance. In this paper, a CDRRA algorithm, which integrated the routing layers and tracks to address the crosstalk noise issue during the track/layer assignment stage, is proposed. The CDRRA problem is formulated as a weighted bipartite matching problem and solved using the linear assignment algorithm. The crosstalk risks between nets are represented by an undirected graph and the maximum number of the concurrent crosstalk risking nets is computed as the max-clique of the graph. Then the nets in each max-clique are assigned to disadjacent tracks. Thus the crosstalk noise can be avoided based on the clique concept. The algorithm is tested by a set of bench mark examples and experimental results show that it can improve the final routing layout a lot with little loss of the completion rate.
{"title":"Crosstalk driven routing resource assignment","authors":"Hailong Yao, Qiang Zhou, Xianlong Hong, Yici Cai","doi":"10.1109/ISCAS.2004.1329465","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1329465","url":null,"abstract":"Crosstalk noise is one of the emerging issues in deep submicrometer technology which causes many undesired effects on the circuit performance. In this paper, a CDRRA algorithm, which integrated the routing layers and tracks to address the crosstalk noise issue during the track/layer assignment stage, is proposed. The CDRRA problem is formulated as a weighted bipartite matching problem and solved using the linear assignment algorithm. The crosstalk risks between nets are represented by an undirected graph and the maximum number of the concurrent crosstalk risking nets is computed as the max-clique of the graph. Then the nets in each max-clique are assigned to disadjacent tracks. Thus the crosstalk noise can be avoided based on the clique concept. The algorithm is tested by a set of bench mark examples and experimental results show that it can improve the final routing layout a lot with little loss of the completion rate.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"26 1","pages":"V-V"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84031303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1328365
Yaxiong Zhang, A. Hamilton
Palmo techniques use pulse width modulated digital signals to convey analog signal information between programmable analog cells. A cell based on CMOS current mode comparator and square root domain integrator circuits is presented. This new implementation of the Palmo cell may be operated in continuous time or sampled data modes to perform a variety of signal processing tasks. Circuits operate at 3V, with sampling rates of up to 5 MHz for sampled data circuits and cut-off frequencies up to 16MHz for continuous time third order low pass filter circuits.
{"title":"Current mode /spl radic/X-domain Palmo technique cell for programmable analog VLSI","authors":"Yaxiong Zhang, A. Hamilton","doi":"10.1109/ISCAS.2004.1328365","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1328365","url":null,"abstract":"Palmo techniques use pulse width modulated digital signals to convey analog signal information between programmable analog cells. A cell based on CMOS current mode comparator and square root domain integrator circuits is presented. This new implementation of the Palmo cell may be operated in continuous time or sampled data modes to perform a variety of signal processing tasks. Circuits operate at 3V, with sampling rates of up to 5 MHz for sampled data circuits and cut-off frequencies up to 16MHz for continuous time third order low pass filter circuits.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"15 1","pages":"I-996"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84033097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1329088
D. Schwingshackl, G. Kubin
This paper proposes a polyphase representation for Volterra filters. To derive the new realizations the well known linear polyphase theory is extended to the nonlinear case starting with Volterra filters. Both the upsampling and downsampling case are considered. As in the linear case (FIR filters) neither the input signal nor the Volterra kernels must fulfill constraints in order to be realized in polyphase form. The computational complexity could be reduced significantly because of two reasons. On the one hand all operations are performed at the low sampling rate and on the other hand some coefficients disappear in the polyphase representation.
{"title":"Polyphase representation of multirate Volterra filters","authors":"D. Schwingshackl, G. Kubin","doi":"10.1109/ISCAS.2004.1329088","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1329088","url":null,"abstract":"This paper proposes a polyphase representation for Volterra filters. To derive the new realizations the well known linear polyphase theory is extended to the nonlinear case starting with Volterra filters. Both the upsampling and downsampling case are considered. As in the linear case (FIR filters) neither the input signal nor the Volterra kernels must fulfill constraints in order to be realized in polyphase form. The computational complexity could be reduced significantly because of two reasons. On the one hand all operations are performed at the low sampling rate and on the other hand some coefficients disappear in the polyphase representation.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"14 1","pages":"IV-653"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78302925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1329695
P. Arena, L. Fortuna, A. Basile, M. Frasca
In this work a methodology for real-time robot navigation in a complex, dynamically changing environment, based on wave computation and implemented by cellular neural networks (CNNs) is introduced. The keypoint of the approach is to consider the environment in which the robot moves as an excitable medium. Obstacles and targets represent the source of autowave generation. The wavefronts propagating in the CNN medium provide to the robot all the information to achieve an adaptive motion avoiding the obstacles and directed to the target. In particular the paradigm of reaction-diffusion (RD) equations are used to implement a CNN-based wave computation for navigation control. Experimental results validating the approach are shown.
{"title":"CNN wave based computation for robot navigation planning","authors":"P. Arena, L. Fortuna, A. Basile, M. Frasca","doi":"10.1109/ISCAS.2004.1329695","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1329695","url":null,"abstract":"In this work a methodology for real-time robot navigation in a complex, dynamically changing environment, based on wave computation and implemented by cellular neural networks (CNNs) is introduced. The keypoint of the approach is to consider the environment in which the robot moves as an excitable medium. Obstacles and targets represent the source of autowave generation. The wavefronts propagating in the CNN medium provide to the robot all the information to achieve an adaptive motion avoiding the obstacles and directed to the target. In particular the paradigm of reaction-diffusion (RD) equations are used to implement a CNN-based wave computation for navigation control. Experimental results validating the approach are shown.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"23 1","pages":"V-V"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78463794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1328959
A. Worapishet, S. Virunphun, M. Chongcheawchamnan, S. Srisathit
An enhanced quadrature LC oscillator based upon the cross coupling between two magnetic-coupled or mutual negative-resistance oscillators is presented. The use of the mutual-R topology results in the entire elimination of the cross-coupling devices normally required when the traditional negative-G/sub m/ oscillator topology is used, thereby yielding significant benefits to power consumption and phase noise performance. Practical simulated results of 3 GHz single and quadrature-phase mutual-R oscillators in a 0.25 /spl mu/m CMOS technology demonstrate the capability of the topology over the conventional negative-G/sub m/ approach.
{"title":"A mutual-negative-resistance quadrature CMOS LC oscillator","authors":"A. Worapishet, S. Virunphun, M. Chongcheawchamnan, S. Srisathit","doi":"10.1109/ISCAS.2004.1328959","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1328959","url":null,"abstract":"An enhanced quadrature LC oscillator based upon the cross coupling between two magnetic-coupled or mutual negative-resistance oscillators is presented. The use of the mutual-R topology results in the entire elimination of the cross-coupling devices normally required when the traditional negative-G/sub m/ oscillator topology is used, thereby yielding significant benefits to power consumption and phase noise performance. Practical simulated results of 3 GHz single and quadrature-phase mutual-R oscillators in a 0.25 /spl mu/m CMOS technology demonstrate the capability of the topology over the conventional negative-G/sub m/ approach.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"9 1","pages":"IV-137"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78019653","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1328988
W. Gee, P. Allen
An active integrated LC filter with tunable gain and quality factor is implemented in a standard 0.18 /spl mu/m CMOS process. The circuit uses transformer feedback and a novel level-shifted single-transistor Q-enhancement technique to maximize the allowable voltage swing across the integrated LC resonator while providing single to differential signal conversion. To illustrate the concept, the circuit is designed and simulated for a wireless receiver front-end. The filter operates at a center frequency of 2.44 GHz with a 3-dB bandwidth of 84 MHz and is targeted at input filtering for Bluetooth and ZigBee applications. Simulation results show that the design achieves a power gain of 4.2 dB, an input 1-dB compression point (P/sub 1dB/) of -4.9 dBm, and a worst-case dynamic range (DR) of 65 dB over the filter bandwidth. Alternate implementations targeting low-voltage and low-power consumption circuits are also presented.
{"title":"CMOS integrated transformer-feedback Q-enhanced LC bandpass filter for wireless receivers","authors":"W. Gee, P. Allen","doi":"10.1109/ISCAS.2004.1328988","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1328988","url":null,"abstract":"An active integrated LC filter with tunable gain and quality factor is implemented in a standard 0.18 /spl mu/m CMOS process. The circuit uses transformer feedback and a novel level-shifted single-transistor Q-enhancement technique to maximize the allowable voltage swing across the integrated LC resonator while providing single to differential signal conversion. To illustrate the concept, the circuit is designed and simulated for a wireless receiver front-end. The filter operates at a center frequency of 2.44 GHz with a 3-dB bandwidth of 84 MHz and is targeted at input filtering for Bluetooth and ZigBee applications. Simulation results show that the design achieves a power gain of 4.2 dB, an input 1-dB compression point (P/sub 1dB/) of -4.9 dBm, and a worst-case dynamic range (DR) of 65 dB over the filter bandwidth. Alternate implementations targeting low-voltage and low-power consumption circuits are also presented.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"37 1","pages":"IV-IV"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73114873","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2004-05-23DOI: 10.1109/ISCAS.2004.1329910
Chin-Teng Lin, Chang-Moun Yeh, Chun-Fei Hsu
Fuzzy neural networks (FNNs) for pattern classification usually use the backpropagation or C-cluster type learning algorithms to learn the parameters of the fuzzy rules and membership functions from the training data. However, such kinds of learning algorithms usually cannot minimize the empirical risk (training error) and expected risk (testing error) simultaneously, and thus cannot reach a good classification performance in the testing phase. To tackle this drawback, a support-vector-based fuzzy neural network classification (SVFNNC) is proposed. The SVFNNC combines the superior classification power of support vector machine (SVM) in high reasoning of FNN in handling uncertainty information. The learning algorithm consists of two learning phases. In the phase 1, the fuzzy rules and membership functions are automatically determined by the clustering principle. In the phase 2, the parameters of FNN are calculated by the SVM with the proposed adaptive fuzzy kernel function. To investigate the effectiveness of the proposed SVFNNC, it is applied to the iris, vehicle and dna datasets. Experimental results show that the proposed SVFNNC can achieve good classification performance with drastically reduced number of fuzzy kernel functions.
{"title":"Fuzzy neural network classification design using support vector machine","authors":"Chin-Teng Lin, Chang-Moun Yeh, Chun-Fei Hsu","doi":"10.1109/ISCAS.2004.1329910","DOIUrl":"https://doi.org/10.1109/ISCAS.2004.1329910","url":null,"abstract":"Fuzzy neural networks (FNNs) for pattern classification usually use the backpropagation or C-cluster type learning algorithms to learn the parameters of the fuzzy rules and membership functions from the training data. However, such kinds of learning algorithms usually cannot minimize the empirical risk (training error) and expected risk (testing error) simultaneously, and thus cannot reach a good classification performance in the testing phase. To tackle this drawback, a support-vector-based fuzzy neural network classification (SVFNNC) is proposed. The SVFNNC combines the superior classification power of support vector machine (SVM) in high reasoning of FNN in handling uncertainty information. The learning algorithm consists of two learning phases. In the phase 1, the fuzzy rules and membership functions are automatically determined by the clustering principle. In the phase 2, the parameters of FNN are calculated by the SVM with the proposed adaptive fuzzy kernel function. To investigate the effectiveness of the proposed SVFNNC, it is applied to the iris, vehicle and dna datasets. Experimental results show that the proposed SVFNNC can achieve good classification performance with drastically reduced number of fuzzy kernel functions.","PeriodicalId":6445,"journal":{"name":"2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512)","volume":"58 1","pages":"V-V"},"PeriodicalIF":0.0,"publicationDate":"2004-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73281042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：