"... In PAGE 12: ...2 Performance and Comparisons This section summarizes the performance of the ECP prototype implementations and shows how it compares against leading software and hardware implementa- tions. Table4 summarizes the performance of the ECP prototypes for the two el- liptic curve algorithms. The results in this table illustrate that the Montgomery method is about 1.... In PAGE 12: ...5/6), as the processing time for additions, squares, and inversions is almost negligible. Table4 also shows that the speedup increases as the digit size increases. The increase is not proportional to the digit size.... ..."

"... In PAGE 12: ...2 Performance and Comparisons This section summarizes the performance of the ECP prototype implementations and shows how it compares against leading software and hardware implementa- tions. Table4 summarizes the performance of the ECP prototypes for the two el- liptic curve algorithms. The results in this table illustrate that the Montgomery method is about 1.... In PAGE 12: ...ethod is about 1.7 times faster than the traditional double-and-add algorithm. One can deduce from Table 1 that this is a direct result of the number of mul- tiplications required by each algorithm ( 10:5=6), as the processing time for additions, squares, and inversions is almost negligible. Table4 also shows that the speedup increases as the digit size increases. The increase is not proportional to the digit size.... ..."

"... In PAGE 11: ... Fremont performs these functions without undue consumption of net- work or host system resources. Table7 summarizes the network characteristics that the current prototype discovers, based on the 8 different Explorer Modules we have implemented. This information is sufficient to provide detailed net- work maps, including topology maps (as illustrated in Figure 2), and tables showing the names and addresses of each host on each network, the local gateways used by each host, etc.... ..."

"... In PAGE 11: ... Fremont performs these functions without undue consumption of net- work or host system resources. Table7 summarizes the network characteristics that the current prototype discovers, based on the 8 different Explorer Modules we have implemented. This information is sufficient to provide detailed net- work maps, including topology maps (as illustrated in Figure 2), and tables showing the names and addresses of each host on each network, the local gateways used by each host, etc.... ..."

"... In PAGE 6: ... Since our FVM implementation in UML allows for two adaptation strategies (MPL control and rate control), we must specify constants for the AIMD rules of both control. For the prototype used in this paper, Table1 shows the baseline settings of our FVM implementation in UML. Unless otherwise specified, these settings are used in our experimental... ..."

"... In PAGE 11: ...he network. These side-effects should be also minimized. 5.1 Functionality of the Implementation Four test scenarios are used in the prototype to show the benefits of context-aware handover as depicted in Table2 . Scenario 1 is based on standard Mobile IP version 6 (MIPv6) [16] handover and it is used as a benchmark to evaluate the efficiency of our context-aware handover scenarios.... ..."

"... In PAGE 14: ... The resulting linear system had a dimension of 1000 1000. In Table1 we show the timings achieved with the Python prototype code of ADDFIT on one processing node of Bunyip, and Tables 2 and 3 show timings for the C/MPI implementation. All presented times are in seconds.... ..."

"... In PAGE 7: ... Technology mapping is performed for the Motorola FPAA [2] library. Table2 shows the macros and the programmable parameter settings used to implement the various functions. The analog part was im- plemented on the Motorola MPAA020 FPAA and digital part on a Xilinx XC4010E FPGA, and the prototype was tested for the desired frequency response and also the following sys- tem performance requirements: detection time a189 40ms, and talk-off a189 10 hits.... ..."