Timescale of interest in traffic measurement for link bandwidth allocation design

Abstract

Consider the link bandwidth allocation for transport of correlated traffic through a queueing system under a maximum allowable delay constraint d/sub max/. We decomposed the traffic into three frequency regions: low-frequency traffic in 0<|/spl omega/|/spl les//spl omega//sub L/, high-frequency traffic in |/spl omega/|/spl ges//spl omega//sub H/ and mid-frequency traffic in /spl omega//sub L/<|/spl omega/|</spl omega//sub H/. The zero-frequency component (DC term) of the traffic provides the average input rate which corresponds to the minimum link bandwidth requirement. Subject to delay constraint d/sub max/, we identify /spl omega//sub /spl lambda//=0.01/spl pi//d/sub max/ and /spl omega//sub H/=2/spl pi//d/sub max/. Hence, the transport of low-frequency traffic exceeds the limit of d/sub max/-constrained buffer capacity; its link bandwidth is essentially captured by its peak rate. In contrast, for the transport of high-frequency traffic the d/sub max/-constrained buffering is most effective and no additional link bandwidth is required. Essentially, the solution of /spl omega//sub L/ and /spl omega//sub H/ plays a role as sampling theory in traffic measurement for buffer capacity design and link bandwidth allocation. Equivalently in the time domain, the timescale of the low-frequency traffic is longer than or equal to 200d/sub max/; the timescale of high-frequency traffic is shorter than or equal to d/sub max/. Since the link bandwidth allocation of low- and high-frequency traffic requires no measurement of second-order statistics, the timescale of interest for traffic measurement must be identified in [d/sub max/, 200d/sub max/].