5.5 Data Analysis of Experiment Three
5.5.1 Analysis in the interval of the object dispatch and the
computation-capability for influence of device operation delay
According to the test results in Analysis 2-2, it indicated that the different communications medium usages would not infulence the device operation noticeably.
Therefore, we have to continue to understand the relationships between these two factors in the operation efficiencies of TCIS devices without the distinctios between the communications medium usages futher. Table 5.15 shows the observed mean of TCIS device operatios in different Cc and Mi of Anaiysis 3-1.
Table 5.15 Observed mean of TCIS device operations (Analysis 3-1)
C
cM
iMean
(ms)Std. Deviation N C
mTCIS-i
2-second 945.3450 166.44662 200
3.5G mobile communications + W ired Network Communications
5-second 920.7750 139.17851 200 10-second 912.5800 137.13051 200 30-second 925.3350 154.82719 200 60-second 924.0200 142.32290 200 Total 925.6110 148.49386 1000
TCIS-ii
2-second 1786.6150 243.19388 200 5-second 1860.4950 245.72702 200 10-second 1766.2200 245.53423 200 30-second 1869.7050 605.03442 200 60-second 1861.2900 262.38932 200 Total 1828.8650 352.61996 1000 TCIS-iii 2-second 2457.4900 473.64275 200
5-second 2433.1000 433.78753 200
10-second 2446.6000 408.93746 200 30-second 2464.1950 354.22174 200 60-second 2507.0350 428.18940 200 Total 2461.6840 421.45668 1000
TCIS-iv
2-second 4466.6150 791.07827 200 5-second 4447.3550 770.67694 200 10-second 4448.0600 735.94570 200 30-second 4479.1100 785.31021 200 60-second 4436.6550 662.16477 200 Total 4455.5590 749.18885 1000
Total
2417.9297 1379.69010 4000As Table 5.15 represcented, the relationships between Cc and Mi can be illustrated as Figure 5.10 and 5.11. We can cleanly know that the operation delay might be infulenced by Cc more, and the lower Cc might operate slower.
Figure 5.10 The represection of the factor in Cc (Analysis 3-1)
Figure 5.11 The represection of the factor in Mi (Analysis 3-1)
Next, Table 5.16 is the analysis results which are analyzed by Two-Way ANOVA method and the alternative conclusions for the test are below:
The alternatives for the analysis of the different computation-capability of TCIS device operations (Cc) are:
H0: The factor in the computation-capability of TCIS devices would not influence the device operation delay
H1: The factor in the computation-capability of TCIS devices would influence the device operation delay
The alternatives for the analysis of the interval of object dispatch (Mi) are:
H0: The interval of object dispatch is not the noticeable factor in device operation delay
H : The interval of object dispatch is the noticeable factor in device operation
delay
The alternatives for the interaction effects between Cc and Mi are:
H0: The Cc and Mi would not possess the noticeable interaction effects in device operation delay
H1: The Cc and Mi would possess the noticeable interaction effects in device operation delay
Table 5.16 Two-Way ANOVA Table of Analysis 3-1
Source
Type III Sum ofSquares
df
Mean SquareF Sig.
Corrected Model 6.731E9 19 3.542E8 1599.341 .000 Intercept 2.339E10 1 2.339E10 105579.705 .000
C
c 6.728E9 3 2.243E9 10124.854 .000M
i 885994.784 4 221498.696 1.000 .406C
c*M
i 1973485.288 12 164457.107 .742 .711 Error 8.816E8 3980 221496.518Total 3.100E10 4000
Corrected Total 7.612E9 3999
a. R Squared = .884 (Adjusted R Squared =.884) b. Computed using alpha = .05
The results represent that the interval of object dispatch (Mi) is not the effective factor in device operations due to the P-value of Mi is 0.406>0.050. On the other hand, because of the P-value of Cc is 0.000<0.050. Therefore, the factor in the computation-capability of TCIS devices might be noticeable factor in the operation delay further, and this result also confirm to Figure 5.10 additionly. Besides, the Mi and
C
c would not have interaction effects between them.5.5.2 Analysis in the interval of the object dispatch and the
computation-capability for influences of data transmission delay
Although the interval of object dispatch Mi is not the noticeable factor in device operations, but it does not mean that these would also not influence the delay in data packet transmission. Hence, Analysis 3-2 is to test the conditions for both of Mi and Cc
in transmission delay.
As mentioned in the results of Experiment One, the noticeable factors in the data packet transmission delay might be the communications medium usages and DSM addition at least. Therefore, the analyses below are discussed on the factors in differences in the communications media(Cm), the interval of object dispatch (Mi) and the compitaion-capability of TCIS devices (Cc) via the muliple regression equation as Table 5.18 and Equation 5.6 shown.
First, the statistic describuations of Analysis 3-2 are listed in Table 5.17 and Figure 5.12 illustrates that the relationships between Mi and Cc, we could know that the Cc
and Mi might be the factors in data packet transmission by this figure.
Table 5.17 Observed mean of data packet transmission (Analysis 3-2)
C
mC
cMean
(ms)Std. Deviation N
3.5G mobile communications
TCIS-i 332.3420 90.55983 500
TCIS-ii 340.4680 102.15671 500 TCIS-iii 367.1920 142.83247 500
TCIS-iv 392.0860 69.62912 500
Total 358.0220 107.27733 2000
Wired network communications
TCIS-i 248.3107 99.61598 500
TCIS-ii 250.9160 89.64893 500
TCIS-iii 280.6913 105.39957 500 TCIS-iv 293.6893 120.25834 500 Total 268.4018 106.02680 2000
Total
TCIS-i 290.3263 104.02031 1000 TCIS-ii 295.6920 105.99110 1000 TCIS-iii 323.9417 132.70940 1000 TCIS-iv 342.8877 109.85616 1000 Total 313.2119 115.67478 4000
Figure 5.12 The represection of the factor in Cc (Analysis 3-2)
Table 5.18 Regression equation analysis of Analysis 3-2
Parameter β
Std. Errort Sig.
Intercept 313.107 3.999 78.300 .000
[Cm=3.5GMobile] 89.620 3.271 27.400 .000
a. Computed using alpha = .05
b. This parameter is set to zero because it is redundant.
Therefore, the multiple regression equation can be expressed by Equation 5.6 and discussed below:
Thus, the P-values of each factor is equal 0.000<0.050, therefore all the factors in this table are the noticeable factors for the data packet transmission as well (i.e., Cm, Cc and
M
f) and the responses are described as following.i. The value 313.107 is the intercept of μˆAnalysis3−2; it indicates the mean of the
distribution of the time-consumption in data transmission at least;
ii. The coefficient of Cm indicates the change in mean of different communications medium usages, it would increase 89.620 ms while the 3.5G mobile communication usage and all the other independent variables remain
constant;
iii. The compilation-capability would also be the noticeable factor in data packet transmission as well, and;
iv. The tight interval of object dispatch might cause more loads for network transmission; therefore, it would reduce 0.702 ms when Mf increases by one second while all the other independent variables remain constant.
5.5.3 Brief summary
So far, we have understood that the factor in computation-capability Cc would not only influence the device operations but also affect the data packet transmission as well. In addition, acccording to Equation 4.2, the time-consumption of data transmission (socket processes) in this research is caculated from the procedures in data packet sealing by transmitter, data packet transmission by Enternet and data packet unsealing by servicers; therefore, the lower computation-capability would cause the more native-effects in sealing and unsealing procedures for the device operations.This explainations not only interprets that the Factor Cc is one of the noticeable factors in data packet transmission but also confirm that why the DSM addition would infulence the data transmission in Equation 5.3 due to the DSM addition adds more loads for core systmes and it infulence the sealing/unsealing operations further.
On other hand, the time-consumption of data packet transmission is explained by Equation 5.3 and 5.6. Equation 5.3 indicates that the DSM addition and Cm are the noticeable factors in data packet transmission under the fixed Mf and Cc factors. If we respectively set up the Mf and Cc in 10-second and TCIS-ii for Equation 5.7, we can get the formala as Equation 5.6.
)
Equation 5.6 accords with the results in Equation 5.3 without the ineffective factor Ms
further. Hence, we can support that the efficiency difference of data packet transmission between the wired network and the 3.5G mobile communications is around 90 ms, and the mean of data packet transmission in wired network is around 258 ms additionlly.
Finally, Table 5.19 shows the analysis results for all the considered factors (i.e., Cm,
KS
l, KPl, Ms, Mf and Cc) from the above analyses.◎:The noticeable and effective factor
As Figure 5.12 shown, the factor in computation-capability Cc is the most important and major factor for the whole DSM operations, it would influence all the operation procedures of DSM. In addition, Figure 5.1 also illustrates that the lower computation-capability device would reach the limits of device operations (or crash) earlier. In addition, due to there are three procedures in data packet transmission in the DSM operations; hence, the network transmission efficiency is another main factor for DSM operations as well. Therefore, as discussed before, we should not only consider or focus on the security level for the security mechanisms but also need to pay more attentions in the efficiencies of both the device operations and network transmissions.