-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathDissertation.out
75 lines (75 loc) · 6.17 KB
/
Dissertation.out
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
\BOOKMARK [0][-]{Doc-Start}{ Acknowledgements}{}% 1
\BOOKMARK [0][-]{chapter.1}{1 Introduction}{}% 2
\BOOKMARK [1][-]{section.1.1}{1.1 Computer networks pervade}{chapter.1}% 3
\BOOKMARK [1][-]{section.1.2}{1.2 Electricity costs for operational networks}{chapter.1}% 4
\BOOKMARK [1][-]{section.1.3}{1.3 The impact of energy inefficiency on operational networks' electricity costs}{chapter.1}% 5
\BOOKMARK [1][-]{section.1.4}{1.4 Prevalent electricity cost reduction techniques}{chapter.1}% 6
\BOOKMARK [2][-]{subsection.1.4.1}{1.4.1 Reducing the amount of energy consumed}{section.1.4}% 7
\BOOKMARK [2][-]{subsection.1.4.2}{1.4.2 Using cheaper electricity - Workload Relocation \(WR\)}{section.1.4}% 8
\BOOKMARK [1][-]{section.1.5}{1.5 Energy efficiency improvement using WR and RP}{chapter.1}% 9
\BOOKMARK [1][-]{section.1.6}{1.6 Our thesis}{chapter.1}% 10
\BOOKMARK [1][-]{section.1.7}{1.7 Contributions}{chapter.1}% 11
\BOOKMARK [1][-]{section.1.8}{1.8 Organization}{chapter.1}% 12
\BOOKMARK [0][-]{chapter.2}{2 Background - Power consumption models in service provider networks}{}% 13
\BOOKMARK [1][-]{section.2.1}{2.1 Networks of geo-diverse data centers - enablers of cloud computing}{chapter.2}% 14
\BOOKMARK [2][-]{subsection.2.1.1}{2.1.1 Structure of a data center}{section.2.1}% 15
\BOOKMARK [2][-]{subsection.2.1.2}{2.1.2 Interconnection of data centers}{section.2.1}% 16
\BOOKMARK [2][-]{subsection.2.1.3}{2.1.3 Handling of client requests}{section.2.1}% 17
\BOOKMARK [2][-]{subsection.2.1.4}{2.1.4 Data center power consumption model}{section.2.1}% 18
\BOOKMARK [1][-]{section.2.2}{2.2 Cellular networks}{chapter.2}% 19
\BOOKMARK [2][-]{subsection.2.2.1}{2.2.1 Structure of a cellular network}{section.2.2}% 20
\BOOKMARK [2][-]{subsection.2.2.2}{2.2.2 Call placement}{section.2.2}% 21
\BOOKMARK [2][-]{subsection.2.2.3}{2.2.3 BTS power consumption model}{section.2.2}% 22
\BOOKMARK [1][-]{section.2.3}{2.3 A comparison of data center and cellular networks}{chapter.2}% 23
\BOOKMARK [1][-]{section.2.4}{2.4 Discussion}{chapter.2}% 24
\BOOKMARK [0][-]{chapter.3}{3 A generalized model for electricity cost optimization}{}% 25
\BOOKMARK [1][-]{section.3.1}{3.1 Optimization problem model}{chapter.3}% 26
\BOOKMARK [1][-]{section.3.2}{3.2 Optimization problem formulation}{chapter.3}% 27
\BOOKMARK [2][-]{subsection.3.2.1}{3.2.1 The objective function}{section.3.2}% 28
\BOOKMARK [2][-]{subsection.3.2.2}{3.2.2 The constraints}{section.3.2}% 29
\BOOKMARK [2][-]{subsection.3.2.3}{3.2.3 Comments on the problem formulation}{section.3.2}% 30
\BOOKMARK [1][-]{section.3.3}{3.3 Summary}{chapter.3}% 31
\BOOKMARK [0][-]{chapter.4}{4 Case study I: geo-diverse data centers}{}% 32
\BOOKMARK [1][-]{section.4.1}{4.1 Prelude}{chapter.4}% 33
\BOOKMARK [1][-]{section.4.2}{4.2 Related work}{chapter.4}% 34
\BOOKMARK [1][-]{section.4.3}{4.3 Instantiating the generalized optimization formulation}{chapter.4}% 35
\BOOKMARK [1][-]{section.4.4}{4.4 Sources of transition costs in the data center scenario}{chapter.4}% 36
\BOOKMARK [2][-]{subsection.4.4.1}{4.4.1 Problem complexity and a heuristic}{section.4.4}% 37
\BOOKMARK [1][-]{section.4.5}{4.5 Experimental setup}{chapter.4}% 38
\BOOKMARK [2][-]{subsection.4.5.1}{4.5.1 Application workload}{section.4.5}% 39
\BOOKMARK [2][-]{subsection.4.5.2}{4.5.2 Electricity prices}{section.4.5}% 40
\BOOKMARK [2][-]{subsection.4.5.3}{4.5.3 Algorithms for workload distribution/relocation}{section.4.5}% 41
\BOOKMARK [1][-]{section.4.6}{4.6 Results}{chapter.4}% 42
\BOOKMARK [2][-]{subsection.4.6.1}{4.6.1 Sensitivity of electricity cost savings to extent of over provisioning}{section.4.6}% 43
\BOOKMARK [2][-]{subsection.4.6.2}{4.6.2 Sensitivity of electricity cost savings to magnitude of transition costs}{section.4.6}% 44
\BOOKMARK [2][-]{subsection.4.6.3}{4.6.3 Sensitivity of electricity cost savings to resource pruning granularity}{section.4.6}% 45
\BOOKMARK [2][-]{subsection.4.6.4}{4.6.4 Sliding window re-optimization}{section.4.6}% 46
\BOOKMARK [2][-]{subsection.4.6.5}{4.6.5 Sensitivity of electricity cost savings to the server idle-peak power ratio}{section.4.6}% 47
\BOOKMARK [2][-]{subsection.4.6.6}{4.6.6 Performance of the heuristic algorithm}{section.4.6}% 48
\BOOKMARK [1][-]{section.4.7}{4.7 Discussion}{chapter.4}% 49
\BOOKMARK [2][-]{subsection.4.7.1}{4.7.1 Impact on clients}{section.4.7}% 50
\BOOKMARK [2][-]{subsection.4.7.2}{4.7.2 Implementation of workload relocation}{section.4.7}% 51
\BOOKMARK [2][-]{subsection.4.7.3}{4.7.3 Diversity in workload}{section.4.7}% 52
\BOOKMARK [2][-]{subsection.4.7.4}{4.7.4 Potential increased network traffic}{section.4.7}% 53
\BOOKMARK [1][-]{section.4.8}{4.8 Summary}{chapter.4}% 54
\BOOKMARK [0][-]{chapter.5}{5 Case Study II: Cellular Networks}{}% 55
\BOOKMARK [1][-]{section.5.1}{5.1 Prelude}{chapter.5}% 56
\BOOKMARK [2][-]{subsection.5.1.1}{5.1.1 Motivating example}{section.5.1}% 57
\BOOKMARK [1][-]{section.5.2}{5.2 Related work}{chapter.5}% 58
\BOOKMARK [1][-]{section.5.3}{5.3 Instantiating the generalized optimization formulation}{chapter.5}% 59
\BOOKMARK [2][-]{subsection.5.3.1}{5.3.1 Multi-BTS cellular setting}{section.5.3}% 60
\BOOKMARK [2][-]{subsection.5.3.2}{5.3.2 Problem complexity}{section.5.3}% 61
\BOOKMARK [2][-]{subsection.5.3.3}{5.3.3 Heuristic solution to RED-BL for cellular networks}{section.5.3}% 62
\BOOKMARK [1][-]{section.5.4}{5.4 Experimental setup}{chapter.5}% 63
\BOOKMARK [2][-]{subsection.5.4.1}{5.4.1 Site characteristics}{section.5.4}% 64
\BOOKMARK [1][-]{section.5.5}{5.5 Results}{chapter.5}% 65
\BOOKMARK [2][-]{subsection.5.5.1}{5.5.1 BTS with two possible power states}{section.5.5}% 66
\BOOKMARK [2][-]{subsection.5.5.2}{5.5.2 Multi-state BTS}{section.5.5}% 67
\BOOKMARK [2][-]{subsection.5.5.3}{5.5.3 Performance of heuristic algorithm}{section.5.5}% 68
\BOOKMARK [2][-]{subsection.5.5.4}{5.5.4 Sensitivity to the value of }{section.5.5}% 69
\BOOKMARK [2][-]{subsection.5.5.5}{5.5.5 Increase in call blocking probability}{section.5.5}% 70
\BOOKMARK [1][-]{section.5.6}{5.6 Limitations of our work}{chapter.5}% 71
\BOOKMARK [1][-]{section.5.7}{5.7 Summary}{chapter.5}% 72
\BOOKMARK [0][-]{chapter.6}{6 Conclusions and future work}{}% 73
\BOOKMARK [1][-]{section.6.1}{6.1 Scope of our work}{chapter.6}% 74
\BOOKMARK [1][-]{section.6.2}{6.2 Future work}{chapter.6}% 75