"... In PAGE 21: ...3.1 EXTERNAL ETHERNET TRAFFIC Table2 summarizes the external Ethernet traffic data analyzed below. We consider the two most recent measurement traces i.... In PAGE 21: ...ata sets JAN90E.LP, JAN90E.MP, and JAN90E.HP in Table2 represent the portions of external packets from the corresponding data sets in Table 1, namely JAN90.... In PAGE 21: ... Repeating the same laborious analysis of Section 4.2 to the data sets described in Table2 , we find that in terms of its self-similar nature, external traffic does not differ from the internal traffic studied earlier. More specifically, the Hurst parameters for the external traffic during normal and high (internal) traffic hours (or during... ..."

"... In PAGE 21: ...3.1 EXTERNAL ETHERNET TRAFFIC Table2 summarizes the external Ethernet traffic data analyzed below. We consider the two most recent measurement traces i.... In PAGE 21: ...ata sets JAN90E.LP, JAN90E.MP, and JAN90E.HP in Table2 represent the portions of external packets from the corresponding data sets in Table 1, namely JAN90.... In PAGE 21: ... Repeating the same laborious analysis of Section 4.2 to the data sets described in Table2 , we find that in terms of its self-similar nature, external traffic does not differ from the internal traffic studied earlier. More specifically, the Hurst parameters for the external traffic during normal and high (internal) traffic hours (or during... ..."

"... In PAGE 21: ...3.1 EXTERNAL ETHERNET TRAFFIC Table2 summarizes the external Ethernet traffic data analyzed below. We consider the two most recent measurement traces i.... In PAGE 21: ...ata sets JAN90E.LP, JAN90E.MP, and JAN90E.HP in Table2 represent the portions of external packets from the corresponding data sets in Table 1, namely JAN90.... In PAGE 21: ... Repeating the same laborious analysis of Section 4.2 to the data sets described in Table2 , we find that in terms of its self-similar nature, external traffic does not differ from the internal traffic studied earlier. More specifically, the Hurst parameters for the external traffic during normal and high (internal) traffic hours (or during... ..."

"... In PAGE 15: ...t 100:1(0.24bpp, 4.24MB) and 60:1(0.4bpp, 7.07MB). The peak signal-to-noise-ratio (PSNR) between the original and decompressed scene is used as the objective measure of the compression quality. We report the PSNRs of all three color components (Y, U and V) in Table1 , however, it is the PSNR result of the Y component that matters most. Therefore, we comment only on the Y component PSNR in the discussion.... In PAGE 22: ...- 22 - Indexes for Tables and Figures Table1 Compression results for the concentric mosaic scenes .... ..."

"... In PAGE 15: ...0:1(0.4bpp, 7.07MB). The peak signal-to-noise-ratio (PSNR) between the original and decompressed scene is used as the objective measure of the compression quality. We report the PSNRs of all three color compo- nents (Y, U and V) in Table1 , however, it is the PSNR result of the Y component that matters most. Therefore, we comment only on the Y component PSNR in the discussion.... In PAGE 21: ...- 21 - Indexes for Tables and Figures Table1 Compression results for the concentric mosaic scenes.... ..."

"... In PAGE 5: ...0:1(0.4bpp, 7.07MB). The peak signal-to-noise-ratio (PSNR) between the original and decompressed scene is shown as the objective measure of compression quality. We report the PSNR of all three color components (Y, U and V) in Table1 , however, it is the PSNR result of the Y component that matters most. Therefore, we comment only on the Y component PSNR in the discussion.... ..."

"... In PAGE 27: ...27 alters, and the density of strong, weak and missing ties between alters. As can be seen from Table3 , the final list included more strong-tie network partners (on average, 79% of contacts, SD=15%). Density of strong and weak ties between alters was similar (.... In PAGE 27: ...25, respectively). Network size, density, constraint and efficiency (Burt, 1992, 1998) were calculated for each of the 125 egocentric networks. Table3 (lower panel) illustrates the central tendencies for these network measures, and reveals that they have unacceptable skewness and kurtosis. Since the assumption of normality underlies the statistical analyses we wish conduct, these variables were transformed: a square root transformation was performed on network density; network efficiency was reflected and a square root transformation was performed, to create a (normally-distributed) measure of network inefficiency.... In PAGE 27: ...19). ******************** Insert Table3 about here ******************** Next, a triad census was compiled based on the nine possible isomorphic classes of strong/weak tie triads, and transformed into triad proportions to allow comparisons across networks. Table 4 exhibits descriptive statistics for the triad proportions.... ..."

"... In PAGE 16: ...imensions is referred to as a dimension map and describes the universe (i.e., gen- eralization state space) for a particular set of attributes. A sample dimension map is shown in Table9 , where the Dim 1, Dim 2, and Dim 3 columns correspond to the hShape; ANYi, hSize; Package; ANYi, and hColour; ANYi paths of the DGG shown in Figure 5. One dimension is associated with each attribute.... ..."

"... In PAGE 22: ...3.2.2001 22 5 Discussion 5.1 Summary of alternative measures Table4 summarises the discussion of chapter 4 and compares the different measures to the criteria presented in chapters 1 and 3.3.... In PAGE 22: ... The measures in between are presented on light grey background. Based on the results of this research ( Table4 ), no single complexity can be considered superior to the others. However, two measures are inferior to the others: the CNCs and the CI.... In PAGE 22: ... The cyclomatic number however, which is equally simple, is more modular than them and has been applied successfully to other problems. The summary in Table4 shows that the CNCs are dominated by the cyclomatic number S in every criteria. Therefore, CNCs should not be considered any further.... ..."

"... In PAGE 3: ... Figure 1 shows the assembly tree for the product. Table2 illustrates a breakdown of op- erations, their descriptions, and the assembly precedence relationships between the operations. 2.... ..."