Abstract :
Network traffic is the outcome of many different sources, which in fact is reminiscent of chaos, that is fractals. Scaling behavior is intrinsically connected with fractals and chaos theory. It is well known that the scaling behavior of teletraffic data affects substantially the overall performance of computer networks. In fact, this behavior is taken into account during analysis, design and deployment of computer networks. Several formulas have been devised to predict queue lengths that conform to desirable QoS (quality of service), given certain traffic characteristics. However, we have not yet found a unified approach that provides good results in every possible case. We explore how elements of chaos theory and fractals can help us design better networks. Networks that provide the QoS they are designed to render.
Keywords :
chaos; computer networks; fractals; quality of service; queueing theory; telecommunication traffic; wavelet transforms; QoS; chaos; chaos theory; computer networks; fractals; network traffic; quality of service; queue length; scaling; teletraffic data; traffic characteristics; wavelets; Aggregates; Computer networks; Feature extraction; Fractals; Gaussian noise; Queueing analysis; Telecommunication traffic; Traffic control; Wavelet analysis; Wavelet coefficients;
