Online Control of Datacenter Power Supply under Uncertain Demand and Renewable Energy
|
|
- Ferdinand Bates
- 8 years ago
- Views:
Transcription
1 Online Control of Datacenter Power Supply under Uncertain Demand and Renewable Energy Wei Deng, Fangming Liu, Hai Jin, Xiaofei Liao Services Computing Technology and System Lab, Cluster and Grid Computing Lab School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan, , China {wdeng, fmliu, hjin, Abstract Modern Cloud Service Providers (CSPs) equip their Datacenter Power Supply System (DPSS) with multisources to mitigate power cost, carbon emission, and power outage: (1) multi-markets grid with time-varying energy prices, (2) finite capacity of Uninterrupted Power Supply (UPS), and (3) certain volumes of intermittent renewable energy. With the presence of uncertain renewable sources and datacenter power demand, CSPs have a critical challenge: how to design systematical online control policies that best utilize different characteristics of multisources in a complementary manner to deliver reliable energy to datacenters while minimizing DPSS operation cost. Based on a stochastic optimization model that captures characteristics of DPSS, we apply two-stage Lyapunov optimization to design and analyze an online DPSS control algorithm (OnDPSS). OnDPSS makes decisions on fully utilizing renewable energy, two-timescales power purchasing, and UPS charging/discharging without requiring substantial statistics of system dynamics. Our mathematical analyses and one-month trace-driven simulations have demonstrated both the optimality (in terms of tradeoff between minimization of DPSS operational cost and constraint satisfaction on datacenter availability and UPS lifetime) and system stability (in terms of robustness to time-varying power demand and supply) achieved by OnDPSS algorithm. I. INTRODUCTION AND RELATED WORK The fast proliferation of the Internet-scale systems and cloud computing services has promoted the massive, geographically distributed datacenters. Large cloud service providers (CSPs) have three major problems during the operation of such datacenters: (1) skyrocketing power consumption and annual electricity bills, such as Google (> 1, 120GW h and > $67M) and Microsoft (> 600GW h and > $36M) [1]. (2) Serious environmental impact. Reportedly, IT carbon footprints occupy 2% of global CO 2 emissions [2]. (3) Unplanned power outages. For instance, Amazon experienced two outages in June 2012 in its US-East-1 region, which were triggered by a series of failures in the power infrastructure [3]. To mitigate such power cost, carbon emission and power outage, modern CSPs usually equip their datacenter power supply systems (DPSS) with multisources in a complementary manner, as illustrated in Fig. 1. First, the grid is still the dominant way to power datacenters for its stable and abundant supply. CSPs can procure power from multiple timescales electricity markets with different energy prices, ranging from long-term market to real-time market [4] [8]. Second, Uninterrupted Power Supply (UPS) is used as the most common way to back up DPSS when there s a grid outage [9]. Recent studies have shown additional benefits of UPSs for reducing power The Corresponding Author is Fangming Liu (fmliu@hust.edu.cn). Uncertain Demand d(t) Long-term ahead planning Datacenter Power Supply System Online Control DPSS to Match Demand Real-time Balancing Supply g(t) ylt(t) yrt(t) Discharge D(β(t)) Long-term Real-time UPS Renewables Multi-peroid Grid Markets Battery Charge R(t) Automatic Transfer Switch(ATS) CSP Datacenters Fig. 1. System model of matching uncertain power demand and supply for CSP to minimize operational cost. cost, by shaving peaks of power demand or shifting demand away from high tariff periods [9], [10]. Third, many CSPs have started to green their datacenters with renewable energy, such as solar and wind, for lower power cost and carbon footprints [2]. In practice, CSPs can automatically switch their power sources via Automatic Transfer Switch (ATS) [11]. When operating such DPSS, there are several key control decisions to be made: (1) How much power to be purchased from the long-term grid market and real-time market, respectively? (2) How to fully leverage the available renewable energy? (3) How to opportunistically use the UPS battery to reduce power cost? With the presence of uncertain power supply and demand, it is challenging to take different characteristics of multiple sources into the consideration of the above decision-making for delivering reliable energy. On the demand side, due to time-varying resource usages of diverse application workloads, a datacenter s power demand is usually variable and uncertain. On the DPSS side, due to the variable energy prices, purchasing power from the grid with multiple markets results in time-varying costs. Furthermore, the intermittent and unpredictable nature of renewable energy aggravates supply-side uncertainty. In addition, the charging/discharging of finite capacity of UPS impacts UPS lifetime and datacenter availability in terms of the continued operation of datacenter [9], [10]. Although many existing works attempted to handle the supply or demand uncertainties in smart grid and datacenters, they either (1) focused on managing the energy supply by assuming a priori knowledge of energy demand [5], [7], [8], [12], or (2) required substantial statistics of the system dynamics with excessive computational complexity [4], [6], or (3) limited to only a single-day optimization of the system [6], or (4) relied on complex forecast techniques of renewable energy [13], [14]. On the contrary, we seek to design online DPSS control policies for a long-term running under uncertain power demand and supply, without requiring substantial
2 system statistics and forecast techniques. Based on a stochastic optimization model that captures time-varying power demand, renewable energy supply, finite UPS and two-timescale grid markets, we derive an online DPSS control algorithm (OnDPSS) by applying two-stage Lyapunov optimization techniques [15], [16]. OnDPSS makes online decisions on fully utilizing renewable energy, twotimescales power purchasing and UPS charging/discharging in a complementary manner to minimize DPSS operation cost. OnDPSS can approach the optimal solution within provable O(1/V ) deviation without requiring substantial statistics of system dynamics. In particular, the control parameter V serves as a control knob that allows CSPs to adjust V that controls the trade-off between minimization of DPSS operation cost and violation of constraints on datacenter availability, UPS lifetime and power capacity. Our mathematical analyses and one-month trace-driven simulations have demonstrated both the optimality (in terms of a well balanced trade-off between DPSS operation cost and constraints violation) and system stability (in terms of robustness and adaptivity to time-varying power demand and supply) achieved by OnDPSS. II. DPSS: OPERATIONAL MODEL AND OBJECTIVE The DPSS system operates in a discrete-time model. Without loss of generality, time is divided into K(K N + ) coarsegrained slots of length T in accordance with the length of longterm ahead market, e.g., days or hours [6]. Each coarse-grained time slot is further divided into N T (N T N + ) fine-grained time slots of length T 1. Empirically, each fine-grained time slot T 1 could be 15 or 60 minutes per which the datacenter can adjust its power control strategies [7], [16]. A. Online Control Decisions As illustrated in Fig. 1, the operation of DPSS includes three key control decisions: 1) Long-term ahead power procurement: At each coarsegrained time slot t = kt (k = 1, 2,..., K), the CSP observes the amount of available renewable energy, denoted as g lt (t), and power demand of a datacenter, denoted as d lt (t). Then, the CSP decides how much amount of power y lt (t) to be purchased from the long-term ahead market at a price p lt (t) for each period τ [t, t + T 1]. 2) Real-time power procurement: At each fined-grained time slot τ [t, t + T 1], the actual power demand d(τ) and available renewable energy g(τ) can be readily observed by the CSP. If the long-term ahead purchasing and the renewable energy are enough to meet the current power demand, i.e., y lt (t) + g(τ) d(τ), then no further procurement is needed. Otherwise, the CSP decides how much additional power y rt (τ) to be purchased from real-time market with a price p rt (τ). 3) Real-time UPS charging/discharging: We assume that CSP can switch between different power sources via automatic transfer switch automatically [11]. If no real-time procurement is needed, then R(τ) = y lt (t) + g(τ) d(τ) will be the excessive energy that can be used to charge the UPS, where R(τ) represents the charged battery energy. Otherwise, the CSP has to make a decision of whether to opportunistically discharge energy D(β(τ)) from the battery along with realtime power procurement to fulfill the current demand: y lt (t) + y rt (τ) + D(β(τ)) + g(τ) = d(τ), (1) where D(β(τ)) denotes the amount of UPS energy discharged at the depth-of-discharge (DoD) level of β(τ) (β(τ) [0, 1]). Let m(t) denote the UPS energy level. We assume that the efficiencies of UPS charging and discharging are the same, denoted by η [0, 1], e.g., η = 0.8 means that only 80% of the charged or discharged energy is useful when charging or discharging. The dynamics of UPS level m(t) can be expressed as below [10], [12]: m(t + 1) = m(t) + ηr(t) D(β(t))/η. (2) B. Online Control Challenges and Constraints There are a series of challenges in decision-making: 1) Power procurement accuracy and cost: Intuitively, the closer to real-time, the CSP has more accurate information of available renewable energy and datacenter power demand, and hence can make more accurate decision on power purchasing [6], [8]. However, it is known that the real-time energy price is more expensive (e.g., p rt (t 1 ) > p lt (t 2 ) if t 1 < t 2 ) [5], [6], [8]. In practice, the price of electricity in real-time markets tends to be higher on average than that in long-term ahead market [5]. The rationale is that upfront payment is associated with cheaper contract prices in the long-term markets. Hence, when procuring power in two-timescale markets (long-term and real time markets), the CSP should make the best tradeoff between procurement accuracy and power cost. 2) Grid capacity limit: We assume that the maximum amount of power that the CSP can draw from the grid in any time period is limited by P grid : 0 y lt (t) + y rt (t) P grid. (3) 3) Datacenter availability constraint: It is extremely critical to guarantee desired datacenter availability with finite capacity of UPS. For instance, ebuff [9] always leaves fiveminutes reserved energy in batteries to ensure datacenter availability. We assume that UPS has a capacity of M UP S, and we choose to store M min minutes of energy to ensure datacenter availability: M min m(t) M UP S. (4) 4) UPS recharge/discharge constraint: In practice, battery has constraints on the maximum amounts of power by which we can recharge or discharge the battery [10]: 0 D(β(t)) D max, 0 R(t) R max. (5) 5) UPS operational cost and lifetime constraint: It has been practically shown that UPS lifetime is a decreasing function of DoD and charge/discharge cycles [17]. The cost of repeated operation of the battery is a function of how often/much it is charged and discharged. Each recharge and discharge operation incurs a cost of C r and C d (β) (at DoD level of β) in dollars, respectively. If a new UPS costs C ups dollars to purchase and it can sustain L ups charge/discharge cycles at maximum DoD, then C r = C ups /L ups, C d (β) =
3 βc ups /L ups. During the long runtime horizon t KT, the maximum allowable discharging number N max satisfies: a(τ) N max, (6) where a(t) = 1 if D(t) > 0, otherwise a(t) = 0. We simplify each charging/discharging cost as a(t)c r (ignoring the impact of DoD β on discharging cost). C. Stochastic Cost Minimization Problem Since the primary costs for renewable energy generation are construction costs such as deploying solar panels and wind turbines, their operational cost is negligible [11]. At time t, the CSP s cost is the power procurement cost plus UPS charging/discharging cost, i.e., Cost(t) y lt (t)p lt (t) + y rt (τ)p rt (τ)) + a(τ)c r. We seek to design an online DPSS control algorithm for solving the following stochastic cost minimization problem: 1 Cost av lim inf E[Cost(τ)](7) s.t. t : constraints (1)(2)(3)(4)(5)(6). min y lt,y rt,d(β),r Lemma 1: In every optimal solution of the optimization problem (7), it holds that y rt (τ) 0, or p rt (t) 0 for τ. Corollary 1: There exits an optimal solution of the above optimization problem (7) wherein y rt (τ) 0 for τ. All complete proofs of Lemma 1, Corollary 1 and Theorems (Sec. III) are referred to our detailed technical report [18]. Remark: The above implies that real-time purchasing is unnecessary (y rt (τ) = 0) in the optimal condition, where all the information about datacenter workloads, renewable energy and energy prices are known in advance. However, this would be too idealized. Without such statistics in practice, how to perform real-time purchasing and battery operation becomes an important problem that we should investigate. To design a flexible and robust online control policy, traditional dynamic programming [6] or Markov decision process [4] require substantial statistics of the system dynamics with excessive computational complexity. In contrast, the recently developed Lyapunov framework is shown to enable the design of online control algorithms for such constrained optimization of time-varying systems without requiring a priori knowledge of the workload and cost statistics [15]. In particular, our above model of the two-timescale power purchase and delivery structure well fits the two-timescale Lyapunov optimization [16], that can enable us to perform two levels of control strategies for two levels of granularity. Therefore, we choose to design our online control algorithm based on two-timescale Lyapunov optimization. III. ONDPSS: ALGORITHM DESIGN AND ANALYSIS A. Lyapunov Optimization Framework To solve the problem (7), we first define a set of virtual queues for the constraints that are captured by our above model, so as to transform constraints satisfaction into queue stability problem. The rationale is that in Lyapunov optimization the constraints are satisfied when their corresponding virtual queues are stable [15]. To ensure DPSS constraints of datacenter availability (4), UPS discharge rate constraint (5), and UPS lifetime (6), we define virtual queues with their update equations as follows: H(t + 1) = max[h(t) m(t), 0] + M min, (8) J(t + 1) = max[j(t) N max, 0] + t a(τ), (9) Z(t + 1) = max[z(t) D max, 0] + D(β(t)). (10) We can note that: H(t + 1) H(t) m(t) + M min. Taking expectations on both sides and summing it up from t = 0 to KT 1, we obtain: E[H(t)] E[h(0)] KT 1 t=0 E[H(t)] + T M min. Dividing both sides by t, and taking t, we can see that H(t)/t 0, and thus the constraint M min m(t) follows. Similarly, we verify that the constraints of discharge rate (5) and UPS lifetime (6) will be ensured as well. Let Q(t) = [m(t), H(t), J(t), Z(t)] be a concatenated vector of the actual and virtual queues. We define a quadratic Lyapunov function as: L(Q(t)) 1 2 [m2 (t) + H 2 (t) + J 2 (t) + Z 2 (t)]. Then, the T -slot conditional Lyapunov drift is defined as: (Q(t)) E[L(Q(t+T )) L(Q(t)) Q(t)]. The drift uses the queueing dynamics to capture how the control policies at each period affect the satisfaction of constraints (2), (4), (5) and (6). Stable drift will ensure the constraints. Following the Lyapunov framework of drift-plus-penalty algorithm [15], our control algorithm is designed to make decisions on y lt (t), y rt (t), D(β(t)) and R(t) to minimize an upper bound on the following drift-plus-penalty term every T slots: (Q(t)) + V E{ Cost(τ) Q(t)}, where the control parameter V is chosen according to the CSP to tune the tradeoff between DPSS cost minimization and constraint satisfaction. A weaker constraint satisfaction can reduce cost, yet can incur adverse effects on datacenter availability. With the presence of dynamic system demand and supply, we approximate future statistics as the current statistics [13], so as to avoid the complexity and substantial statistics requirement of forecasts, i.e., m(τ) = m(t), ĝ lt (τ) = g(t), d lt (τ) = d(t) and p rt (τ) = p rt (t) for t < τ t + T 1. Then, the following Theorem 1 gives the analytical bound on the decisions of power purchasing and UPS charging/discharging. Theorem 1: (Drift-plus-Penalty Bound) Considering the Lyapunov function L(Q(t)), to ensure the constraints in (1) (6), the drift-plus-penalty expression satisfies: T (Q(t)) + V E{ Cost(τ) Q(t)} (11) BT + +V E{ Cost(τ) Q(t)} m(τ)[r(τ)η D(β(τ))/η] Q(t)} H(τ)[M min m(τ)] Q(t)} J(τ)[ a(τ) N max] Q(t)} Z(τ)[D(β(τ)) D max Q(t)]}, where B = R 2 maxη 2 + D2 max η 2 + M 2 ups M 2 min 2 + N 2 max + D 2 max.
4 Remark: To minimize the right-hand-side (RHS) of (11), we design our online DPSS control algorithm (OnDPSS) as Algorithm 1. Specifically, OnDPSS is computationally efficient: each time it only has to solve a linear program with four variables (y lt (t), y rt (t), D(β(t)), R(t)) and six linear constraints in Eq. (1)(2)(4)(5)(6) without a priori statistical knowledge about system dynamics. We do not explicitly take the constraint (3) into consideration, as it is easy to automatically ensure grid capacity limit (3) and battery charge rate R(t) [0, R max ]. OnDPSS intends to make full utilization of renewable energy and battery to minimize power procurement. B. Performance Bound and Robustness Analysis In this section, we analyze the performance of OnDPSS algorithm in terms of the gap between the result achieved by OnDPSS and the theoretical optimal solution φ opt of cost minimization problem (7). Theorem 2: (Performance Bound): The average cost Cost av achieved by OnDPSS algorithm satisfies the following bound with any fixed control parameter V (V > 0) and longterm time slot T : 1 OnDP SS Costav lim inf E[Cost(τ)] φopt + B V, where B is given by Theorem 1. Remark: OnDPSS can approach the optimal solution of problem (7) within a deviation of B/V. CSPs can achieve an elastic cost by adjusting the control knob V, according to a desired tradeoff between DPSS cost minimization and constraints satisfaction. The time slot T decides how frequently OnDPSS performs actions of power procurement and battery charging/discharging. Interestingly, the dynamic UPS energy level reflects variable power demand and uncertain renewable energy supply. Since OnDPSS algorithm approximates future battery energy level as its current level, is it robust for OnDPSS algorithm to make its decisions based on the approximated UPS battery level m(t) that is different from the actual UPS capacity, due to uncertainties of datacenter power demand and supply? The following Theorem 3 demonstrates the robustness of OnDPSS. Theorem 3: (OnDPSS Robustness): We assume that the approximated UPS battery level m(t) and its actual level m(t) satisfy m(t) m(t) θ. Then, under OnDPSS algorithm: OnDP SS 1 Costav lim inf E[Cost(τ)] φopt + B ε V, where B is given by Theorem 1, and B ε = B + T θ(d max + R max + M ups + M min ). Here, D max and R max are the maximum amounts of UPS energy recharging and discharging, respectively; M ups and M min are the maximum and minimum UPS energy levels, respectively. Remark: By comparing Theorem 2 and Theorem 3, we observe that V varies positively along with the uncertainties (θ) of datacenter power demand and renewable energy. However, the satisfaction of constraints varies inversely versus these uncertainties. This implies that the robustness of OnDPSS is achieved at the expense of a tradeoff between operational cost minimization and constraints satisfaction. Algorithm 1: Online Control Algorithm for DPSS Operational Cost Minimization (OnDPSS). 0.8) Long-term ahead planing: At each time t = kt (k Z + ), observing available renewable energy g(t), datacenter power demand d(t), and UPS energy level m(t), CSP decides the optimal power procurement in the long-term market y lt (t) to minimize the expected cost as below: min V y lt (t)p lt (t) s.t. y lt (t) + g(t) d(t) + m(t), m(t) M UP S, 0 y lt (t) P grid. 2) Real-time balancing: At real-time τ [t, t + T 1], CSP performs real-time market procurement y rt(τ), and UPS discharging D(β(τ)) and charging R(τ), to minimize the RHS of (11): min V [y rt(τ)p rt(τ)) + a(τ)cr] + m(t)[r(τ) D(β(τ))] H(t)m(t) + J(t) a(τ) + Z(t)D(β(τ)) s.t. (4)(5)(6), y lt (t) + y rt(τ) + g(τ) + D(β(τ)) = d(τ), if y lt (t) + g(τ) > d(τ), R(τ) = y lt (t) + g(τ) d(τ). 3) Queue Update: Update queues using Eq. (2) (8) (9) (10). IV. ONDPSS: PERFORMANCE EVALUATION A. Real-World Traces and Experimental Setup First, to simulate the intermittent availability of renewable energy, we use solar energy data from the Measurement and Instrumentation Data Center (MIDC) [19]. Specifically, we use the meteorological data from Jan. 1st, 2012 to Jan. 31th, 2012 from central U.S.. Second, to simulate the variable electricity prices, we use the price traces from New York Independent System Operator (NYISO) [20]. Third, we use citizens electricity load data from NYISO to simulate the power demand of the CSP. Both traces are also from Jan. 1st, 2012 to Jan. 31th, According to the recent empirical experiments, we assume that the limits of UPS charging/discharging rates are D max = R max = 0.5MW, and charging/discharging costs are C r = C d = 0.1 dollars [10]. The minimum battery level M min is 5 minutes energy of UPS [9]. The maximum number of UPS charge/discharge cycles is L UP S = 5, 000 with 4-year lifetime constraint [9]. The efficiency of UPS charging/discharging is η = 0.8 [12]. We set the grid limit as P grid = 2MW [10]. B. Experimental Results We conduct sensitivity analysis on critical factors to characterize their impact on DPSS operational cost. 1) Impact of Control Parameter V : As shown in Fig. 2, to simulate the day-ahead power market, we fix T to be 24 time slots (one day). We conduct experiments with different values of V, which show that when V increases from 0.5 to 100, OnDPSS achieves a time-average cost that is closer to the optimal solution. This quantitatively confirms Theorem 2 that OnDPSS can approach the optimal solution within a diminishing gap of O(1/V ). This enables CSP to obtain an elastic performance bound by adjusting V within a desired weighted sum of cost and Lyapunov drift of constraints.
5 Day-Average Cost ($) OnDPSS Optimal V (T=24) Fig. 2. Impact of parameter V. Day-Average Cost ($) OnDPSS 1000 Optimal T (V=10) Fig. 3. Impact of time slot T. 2) Impact of Long-term Time Slot T : In Fig. 3, we fix V to be 10 and vary T from 3 time slots (3 hours) to 144 time slots (6 days), which is a sufficient range for exploring the characteristics of different timescales of long-term market. We observe that changing T has relatively less impact on the cost of operating DPSS. The fluctuation of the time-average cost is more notable when T becomes longer. This is because that the term B ε in Theorem 3 is proportional to T, which means the uncertainties of power demand and renewable energy increase with the increasing of T. As uncertainties can be alleviated by battery and two timescales power procurement, the timeaverage cost only fluctuates within [ 1.65%, +2.23%]. This corroborates Theorem 3 that, even with infrequent actions of DPSS operation, OnDPSS can save significant cost. 3) Impact of Battery Capacity: In Fig. 4, we compare the total average cost under different battery sizes (M UP S {0, 0.25, 0.5, 1}MW h) under two-timescale markets over the 31-days period with V = 10 and T = 24. It shows that UPS can bring in significant benefit, i.e., the total average cost decreases with the increasing of battery size. The rationale is that the UPS can offer additional free power charged from superfluous renewable energy to serve the loads, resulting in less purchasing costs. This verifies that the optimal solution is limited mainly by the battery capacity. 4) Impact of Grid Markets: In Fig. 4, we compare the total average cost over a month of three representative cases: two-timescale markets with 0.5M W h UPS, real-time market with 0.5M W h UPS, and real-time market without UPS. Although purchasing power from the real-time market has higher electricity price, the additional real-time procurement can help reduce 4.87% cost. The rationale is that CSP can purchase the exact amount of power from the two-timescale markets to avoid costs due to inaccurate procurement (extra actions of purchasing and charging/discharging). In addition, we can see that the benefit brought by energy storage is about 19.3% higher than that of the markets. In sum, with twotimescale power markets and different battery sizes, OnDPSS can reduce operational cost by 4.87% 24.17%. V. CONCLUSION This paper applied two-stage Lyapunov optimization to design an online control algorithm (OnDPSS) that best utilizes different characteristics of multiple energy sources in a complementary manner to deliver reliable energy to datacenters while minimizing DPSS operational cost, without requiring substantial statistics of system dynamics. Both mathematical analyses and trace-driven evaluations demonstrated the optimality and stability of OnDPSS. It can approach the optimal solution within a diminishing gap of O(1/V ), which is mainly Fig. 4. Total Cost ($) TwoMarket+NoUPS TwoMarket+0.25MWh UPS TwoMarket+0.5MWh UPS TwoMarket+1MWh UPS RealTime+NoUPS RealTime+0.5MWh UPS UPS Size Impact Grid Markets Impact Impact of battery capacity and grid markets structure in total cost. limited by the UPS battery size, multi-timescale grid markets and DPSS operation frequency. VI. ACKNOWLEDGEMENT The research was supported in part by a grant from The National Natural Science Foundation of China (NSFC) under grant No and No , by a grant from the Research Fund of Young Scholars for the Doctoral Program of Higher Education, Ministry of Education, China, under grant No , by the CHUTIAN Scholar Project of Hubei Province. REFERENCES [1] A. Qureshi, Power-Demand Routing in Massive Geo-Distributed Systems, Ph.D. dissertation, Massachusetts Institute of Technology, [2] How Clean is Your Cloud, [3] [4] M. He, S. Murugesan, and J. Zhang, Multiple Timescale Dispatch and Scheduling for Stochastic Reliability in Smart Grids with Wind Generation Integration, in Proc. of INFOCOM, Apr [5] S. Adlakha and A. Wierman, Energy Procurement Strategies in the Presence of Intermittent Sources, Technical report, [6] L. Jiang and S. Low, Multi-period Optimal Procurement and Demand Responses in the Presence of Uncrtain Supply, in Proc. of IEEE Conference on Decision and Control (CDC), [7] L. Huang, J. Walrand, and K. Ramchandran, Optimal Power Procurement and Demand Response with Quality-of-usage Guarantees, Arxiv preprint arxiv: , [8] P. Varaiya, F. Wu, and J. Bialek, Smart Operation of Smart Grid: Risklimiting Dispatch, in Proc. of the IEEE, vol. 99(1), pp , [9] S. Govindan, A. Sivasubramaniam, and B. Urgaonkar, Benefits and Limitations of Tapping into Stored Energy for Datacenters, in Proc. of ACM ISCA, Jun [10] R. Urgaonkar, B. Urgaonkar, M. Neely, and A. Sivasubramanian, Optimal Power Cost Management Using Stored Energy in Data Centers, in Proc. of ACM SIGMETRICS, Jun [11] C. Stewart and K. Shen, Some joules are more precious than others: Managing renewable energy in the datacenter, in Proc. of the Workshop on Power Aware Computing and Systems, [12] L. Huang, J. Walrand, and K. Ramchandran, Optimal Demand Response with Energy Storage Management, Arxiv preprint arxiv: , [13] Í. Goiri, K. Le, T. Nguyen, J. Guitart, J. Torres, and R. Bianchini, GreenHadoop: Leveraging Green Energy in Data-Processing Frameworks, in Proc. of EuroSys, Apr [14] Y. Zhang, Y. Wang, and X. Wang, GreenWare: Greening Cloud-Scale Data Centers to Maximize the Use of Renewable Energy, in Proc. of ACM Middleware, Dec [15] L. Georgiadis, M. Neely, M. Neely, and L. Tassiulas, Resource Allocation and Cross-layer Control in Wireless Networks. Now Pub, [16] Y. Yao, L. Huang, A. Sharma, L. Golubchik, M. Neely et al., Data Centers Power Reduction: A Two Time Scale Approach for Delay Tolerant Workloads, in Proc. of IEEE INFOCOM, Mar [17] FAQ.htm. [18] W. Deng, F. Liu, H. Jin, and X. Liao, Online Control of Datacenter Power Supply under Uncertain Demand and Renewable Energy, HUST Technical report, Sep [19] MIDC, [20] NYISO,
SmartDPSS: Cost-Minimizing Multi-source Power Supply for Datacenters with Arbitrary Demand
SmartDPSS: Cost-Minimizing Multi-source Power Supply for Datacenters with Arbitrary Demand Wei Deng, Fangming Liu, Hai Jin, Chuan Wu 2 Services Computing echnology and System Lab, Cluster and Grid Computing
More informationImpact of workload and renewable prediction on the value of geographical workload management. Arizona State University
Impact of workload and renewable prediction on the value of geographical workload management Zahra Abbasi, Madhurima Pore, and Sandeep Gupta Arizona State University Funded in parts by NSF CNS grants and
More informationData Center Energy Cost Minimization: a Spatio-Temporal Scheduling Approach
23 Proceedings IEEE INFOCOM Data Center Energy Cost Minimization: a Spatio-Temporal Scheduling Approach Jianying Luo Dept. of Electrical Engineering Stanford University jyluo@stanford.edu Lei Rao, Xue
More informationOnline Resource Management for Data Center with Energy Capping
Online Resource Management for Data Center with Energy Capping A. S. M. Hasan Mahmud Florida International University Shaolei Ren Florida International University Abstract The past few years have been
More informationProfit Maximization and Power Management of Green Data Centers Supporting Multiple SLAs
Profit Maximization and Power Management of Green Data Centers Supporting Multiple SLAs Mahdi Ghamkhari and Hamed Mohsenian-Rad Department of Electrical Engineering University of California at Riverside,
More informationOnline Resource Management for Data Center with Energy Capping
Online Resource Management for Data Center with Energy Capping Hasan Mahmud and Shaolei Ren Florida International University 1 A massive data center Facebook's data center in Prineville, OR 2 Three pieces
More informationThe Answer Is Blowing in the Wind: Analysis of Powering Internet Data Centers with Wind Energy
The Answer Is Blowing in the Wind: Analysis of Powering Internet Data Centers with Wind Energy Yan Gao Accenture Technology Labs Zheng Zeng Apple Inc. Xue Liu McGill University P. R. Kumar Texas A&M University
More informationCutting Down Electricity Cost in Internet Data Centers by Using Energy Storage
his full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE Globecom 0 proceedings. Cutting Down Electricity Cost in Internet
More informationDynamic Virtual Machine Allocation in Cloud Server Facility Systems with Renewable Energy Sources
Dynamic Virtual Machine Allocation in Cloud Server Facility Systems with Renewable Energy Sources Dimitris Hatzopoulos University of Thessaly, Greece Iordanis Koutsopoulos Athens University of Economics
More informationData Center Optimization Methodology to Maximize the Usage of Locally Produced Renewable Energy
Data Center Optimization Methodology to Maximize the Usage of Locally Produced Renewable Energy Tudor Cioara, Ionut Anghel, Marcel Antal, Sebastian Crisan, Ioan Salomie Computer Science Department Technical
More informationENERGY EFFICIENT AND REDUCTION OF POWER COST IN GEOGRAPHICALLY DISTRIBUTED DATA CARDS
ENERGY EFFICIENT AND REDUCTION OF POWER COST IN GEOGRAPHICALLY DISTRIBUTED DATA CARDS M Vishnu Kumar 1, E Vanitha 2 1 PG Student, 2 Assistant Professor, Department of Computer Science and Engineering,
More informationLeveraging Renewable Energy in Data Centers: Present and Future
Leveraging Renewable Energy in Data Centers: Present and Future Ricardo Bianchini Department of Computer Science Collaborators: Josep L. Berral, Inigo Goiri, Jordi Guitart, Md. Haque, William Katsak, Kien
More informationProvably-Efficient Job Scheduling for Energy and Fairness in Geographically Distributed Data Centers
3nd IEEE International Conference on Distributed Computing Systems Provably-Efficient Job Scheduling for Energy and Fairness in Geographically Distributed Data Centers Shaolei Ren Yuxiong He Fei Xu Electrical
More informationOptimal Power Cost Management Using Stored Energy in Data Centers
1 Optimal Power Cost Management Using Stored Energy in Data Centers Rahul Urgaonkar, Bhuvan Urgaonkar, Michael J. Neely, and Anand Sivasubramaniam Dept. of EE-Systems, University of Southern California
More informationEnergy Management and Profit Maximization of Green Data Centers
Energy Management and Profit Maximization of Green Data Centers Seyed Mahdi Ghamkhari, B.S. A Thesis Submitted to Graduate Faculty of Texas Tech University in Partial Fulfilment of the Requirements for
More informationetime: Energy-Efficient Transmission between Cloud and Mobile Devices
eime: Energy-Efficient ransmission between Cloud and Mobile Devices Peng Shu 1,2 Fangming Liu 1,2 Hai Jin 1,2 Min Chen 2 Feng Wen 1,2 Yupeng Qu 1,2 Bo Li 3 1 Key Laboratory of Services Computing echnology
More informationData Centers to Offer Ancillary Services
Data Centers to Offer Ancillary Services Mahdi Ghamkhari and Hamed Mohsenian-Rad Department of Electrical Engineering, University of California at Riverside, Riverside, CA, USA e-mails: {ghamkhari, hamed}@ee.ucr.edu
More informationCost-Minimizing Preemptive Scheduling of MapReduce Workloads on Hybrid Clouds
Cost-Minimizing Preemptive Scheduling of MapReduce Workloads on Hybrid Clouds Xuanjia Qiu, Wai Leong Yeow, Chuan Wu, Francis C.M. Lau Department of Computer Science, The University of Hong Kong, Hong Kong,
More informationCarbon-aware Load Balancing for Geo-distributed Cloud Services
Carbon-aware Load Balancing for Geo-distributed Cloud Services Zhi Zhou 1 Fangg Liu 1 Yong Xu 1 Ruolan Zou 1 Hong Xu 2 John C.S. Lui 3 Hai Jin 1 1 Key Laboratory of Services Computing echnology and System,
More informationEnergy Trading in the Smart Grid: From End-user s Perspective
Energy Trading in the Smart Grid: From End-user s Perspective Shengbo Chen, Ness B. Shroff and Prasun Sinha Department of ECE, The Ohio State University Department of CSE, The Ohio State University Email:
More informationData Centers Power Reduction: A two Time Scale Approach for Delay Tolerant Workloads
Data Centers Power Reduction: A two Time Scale Approach for Delay Tolerant Workloads Yuan Yao, Longbo Huang, Abhihshek Sharma, Leana Golubchik and Michael Neely University of Southern California, Los Angeles,
More informationCarbon-aware Load Balancing for Geo-distributed Cloud Services
Carbon-aware Load Balancing for Geo-distributed Cloud Services Zhi Zhou Fangg Liu Yong Xu Ruolan Zou Hong Xu 2 John C.S. Lui 3 Hai Jin Key Laboratory of Services Computing echnology and System, Ministry
More informationMinimizing the Operational Cost of Data Centers via Geographical Electricity Price Diversity
203 IEEE Sixth International Conference on Cloud Computing Minimizing the Operational Cost of Data Centers via Geographical Electricity Price Diversity Zichuan Xu Weifa Liang Research School of Computer
More informationAlgorithms for sustainable data centers
Algorithms for sustainable data centers Adam Wierman (Caltech) Minghong Lin (Caltech) Zhenhua Liu (Caltech) Lachlan Andrew (Swinburne) and many others IT is an energy hog The electricity use of data centers
More informationChange Management in Enterprise IT Systems: Process Modeling and Capacity-optimal Scheduling
Change Management in Enterprise IT Systems: Process Modeling and Capacity-optimal Scheduling Praveen K. Muthusamy, Koushik Kar, Sambit Sahu, Prashant Pradhan and Saswati Sarkar Rensselaer Polytechnic Institute
More informationCutting Down the Energy Cost of Geographically Distributed Cloud Data Centers
Cutting Down the Energy Cost of Geographically Distributed Cloud Data Centers Huseyin Guler 1, B. Barla Cambazoglu 2 and Oznur Ozkasap 1 1 Koc University, Istanbul, Turkey 2 Yahoo! Research, Barcelona,
More informationOn the Cost-QoE Trade-off for Cloud-based Video Streaming under Amazon EC2 s Pricing Models
On the Cost-QoE Trade-off for Cloud-based Video Streaming under Amazon EC2 s Pricing Models Jian He, Yonggang Wen, Member, IEEE, Jianwei Huang, Senior Member, IEEE, Di Wu, Member, IEEE Abstract The emergence
More informationLoad Balancing and Switch Scheduling
EE384Y Project Final Report Load Balancing and Switch Scheduling Xiangheng Liu Department of Electrical Engineering Stanford University, Stanford CA 94305 Email: liuxh@systems.stanford.edu Abstract Load
More informationGreenWare: Greening Cloud-Scale Data Centers to Maximize the Use of Renewable Energy
GreenWare: Greening Cloud-Scale Data Centers to Maximize the Use of Renewable Energy Yanwei Zhang 1, Yefu Wang 1, and Xiaorui Wang 1,2 {yzhang82, ywang38}@eecs.utk.edu xwang@ece.osu.edu 1 Department of
More informationLeveraging Thermal Storage to Cut the Electricity Bill for Datacenter Cooling
Leveraging Thermal Storage to Cut the Electricity Bill for Datacenter Cooling Yefu Wang1, Xiaorui Wang1,2, and Yanwei Zhang1 ABSTRACT The Ohio State University 14 1 1 8 6 4 9 8 Time (1 minuts) 7 6 4 3
More informationEnergy Efficient Geographical Load Balancing via Dynamic Deferral of Workload
2012 IEEE Fifth International Conference on Cloud Computing Energy Efficient Geographical Load Balancing via Dynamic Deferral of Workload Muhammad Abdullah Adnan, Ryo Sugihara and Rajesh K. Gupta University
More informationROUTING ALGORITHM BASED COST MINIMIZATION FOR BIG DATA PROCESSING
ROUTING ALGORITHM BASED COST MINIMIZATION FOR BIG DATA PROCESSING D.Vinotha,PG Scholar,Department of CSE,RVS Technical Campus,vinothacse54@gmail.com Dr.Y.Baby Kalpana, Head of the Department, Department
More informationCyber-Physical Systems: Some Food for Thought
Cyber-Physical Systems: Some Food for Thought Ness B. Shroff Electrical and Computer Engineering & Computer Science and Engineering E-mail: shroff.11@osu.edu What is CPS? By NSF: engineered systems that
More informationMILP Model And Models For Cloud Network Backup
Minimizing Disaster Backup Window for Geo-Distributed Multi-Datacenter Cloud Systems Jingjing Yao, Ping Lu, Zuqing Zhu School of Information Science and Technology University of Science and Technology
More informationOn the Operation and Value of Storage in Consumer Demand Response
On the Operation and Value of Storage in Consumer Demand Response Yunjian Xu and Lang Tong Abstract We study the optimal operation and economic value of energy storage operated by a consumer who faces
More informationOptimal Integration of Renewable Energy Resources in Data Centers with Behind-the-Meter Renewable Generator
1 Optimal Integration of Renewable Energy Resources in Data Centers with Behind-the-Meter Renewable Generator Mahdi Ghamkhari and Hamed Mohsenian-Rad Department of Electrical and Computer Engineering Texas
More informationSIMPLIFIED PERFORMANCE MODEL FOR HYBRID WIND DIESEL SYSTEMS. J. F. MANWELL, J. G. McGOWAN and U. ABDULWAHID
SIMPLIFIED PERFORMANCE MODEL FOR HYBRID WIND DIESEL SYSTEMS J. F. MANWELL, J. G. McGOWAN and U. ABDULWAHID Renewable Energy Laboratory Department of Mechanical and Industrial Engineering University of
More informationEvolution of the smart grid in China
18 Evolution of the smart grid in China Development of this enormous market could shape the future of the smart grid globally. David Xu, Michael Wang, Claudia Wu, and Kevin Chan China has become the world
More informationValue of storage in providing balancing services for electricity generation systems with high wind penetration
Journal of Power Sources 162 (2006) 949 953 Short communication Value of storage in providing balancing services for electricity generation systems with high wind penetration Mary Black, Goran Strbac 1
More informationTraffic-Aware Resource Provisioning for Distributed Clouds
ENERGY EFFICIENCY Traffic-Aware Resource Provisioning for Distributed Clouds Dan Xu, AT&T Labs Xin Liu, University of California, Davis Athanasios V. Vasilakos, Lulea University of Technology, Sweden Examining
More informationInternational Journal of Computer Sciences and Engineering Open Access. Hybrid Approach to Round Robin and Priority Based Scheduling Algorithm
International Journal of Computer Sciences and Engineering Open Access Review Paper Volume-4, Issue-2 E-ISSN: 2347-2693 Hybrid Approach to Round Robin and Priority Based Scheduling Algorithm Garima Malik
More informationDynamic Scheduling and Pricing in Wireless Cloud Computing
Dynamic Scheduling and Pricing in Wireless Cloud Computing R.Saranya 1, G.Indra 2, Kalaivani.A 3 Assistant Professor, Dept. of CSE., R.M.K.College of Engineering and Technology, Puduvoyal, Chennai, India.
More informationA Sequential Game Perspective and Optimization of the Smart Grid with Distributed Data Centers
A Sequential Game Perspective and Optimization of the Smart Grid with Distributed Data Centers Yanzhi Wang, Xue Lin, and Massoud Pedram Department of Electrical Engineering University of Southern California
More informationWhen to Refinance Mortgage Loans in a Stochastic Interest Rate Environment
When to Refinance Mortgage Loans in a Stochastic Interest Rate Environment Siwei Gan, Jin Zheng, Xiaoxia Feng, and Dejun Xie Abstract Refinancing refers to the replacement of an existing debt obligation
More informationRisk-limiting dispatch for the smart grid: some research problems
Risk-limiting dispatch for the smart grid: some research problems Janusz Bialek, University of Edinburgh Pravin Varaiya, University of California, Berkeley Felix Wu, University of Hong Kong 1 Outline 1.
More informationHow To Optimize A Power Plant
Optimizing Data Centres Operation to Provide Ancillary Services On-demand Marcel Antal, Claudia Pop, Dan Valea, Tudor Cioara, Ionut Anghel, Ioan Salomie Technical University of Cluj-Napoca, Cluj-Napoca,
More informationTHE expansion of the Internet has fueled a rise in cloud
IEEE TRANSACTIONS ON SMART GRID, VOL. XX, NO. YY, MONTH YEAR Inter-Datacenter Job Routing and Scheduling with Variable Costs and Deadlines Carlee Joe-Wong, Student Member, IEEE, Ioannis Kamitsos, Sangtae
More informationDynamic Scaling of VoD Services into Hybrid Clouds with Cost Minimization and QoS Guarantee
Dynamic Scaling of VoD Services into Hybrid Clouds with Cost Minimiation and QoS Guarantee Xuanjia Qiu, Hongxing Li, Chuan Wu, Zongpeng Li and Francis C.M. Lau Department of Computer Science, The University
More informationAn Introduction to Variable-Energy-Resource Integration Analysis Energy Exemplar October 2011
An Introduction to Variable-Energy-Resource Integration Analysis Energy Exemplar October 2011 1. Introduction Increased Renewable Portfolio Standards (RPS) are being enacted at the state, provincial, and
More informationModeling the Aggregator problem: the economic dispatch and dynamic scheduling of flexible electrical loads
Modeling the Aggregator problem: the economic dispatch and dynamic scheduling of flexible electrical loads Anna Scaglione June 23, 2014 1 / 51 Premise Balancing the Grid 2 / 51 How can generators know
More informationPreparing for Distributed Energy Resources
Preparing for Distributed Energy Resources Executive summary Many utilities are turning to Smart Grid solutions such as distributed energy resources (DERs) small-scale renewable energy sources and energy
More informationDistributed Control of Heating, Ventilation and Air Conditioning Systems in Smart Buildings
Distributed Control of Heating, Ventilation and Air Conditioning Systems in Smart Buildings Najmeh Forouzandehmehr Electrical and Computer Engineering Department University of Houston Houston, Texas, USA
More informationMulti-service Load Balancing in a Heterogeneous Network with Vertical Handover
1 Multi-service Load Balancing in a Heterogeneous Network with Vertical Handover Jie Xu, Member, IEEE, Yuming Jiang, Member, IEEE, and Andrew Perkis, Member, IEEE Abstract In this paper we investigate
More informationMINIMIZING STORAGE COST IN CLOUD COMPUTING ENVIRONMENT
MINIMIZING STORAGE COST IN CLOUD COMPUTING ENVIRONMENT 1 SARIKA K B, 2 S SUBASREE 1 Department of Computer Science, Nehru College of Engineering and Research Centre, Thrissur, Kerala 2 Professor and Head,
More informationSocially-Responsible Load Scheduling Algorithms for Sustainable Data Centers over Smart Grid
Socially-Responsible Load Scheduling Algorithms for Sustainable Data Centers over Smart Grid Jian He, Xiang Deng, Dan Wu, Yonggang Wen, Di Wu Department of Computer Science, Sun Yat-Sen University, Guangzhou,
More informationGreen-Aware Workload Scheduling in Geographically Distributed Data Centers
Green-Aware Workload Scheduling in Geographically Distributed Data Centers Changbing Chen, Bingsheng He, Xueyan Tang Nanyang Technological University, Singapore 639798 {chchangb, bshe, asxytang}@ntu.edu.sg
More informationCapacity management of vehicle-to-grid system for power regulation services
Title Capacity management of vehicle-to-grid system for power regulation services Author(s) Lam, AYS; Leung, KC; Li, VOK Citation The IEEE 3rd International Conference on Smart Grid Communications (SmartGridComm
More informationOptimize the Dynamic Provisioning and Request Dispatching in Distributed Memory Cache Services
Optimize the Dynamic Provisioning and Request Dispatching in Distributed Memory Cache Services Boyang Yu and Jianping Pan University of Victoria, British Columbia, Canada Abstract The dynamic provisioning
More information[Sathish Kumar, 4(3): March, 2015] ISSN: 2277-9655 Scientific Journal Impact Factor: 3.449 (ISRA), Impact Factor: 2.114
IJESRT INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY HANDLING HEAVY-TAILED TRAFFIC IN QUEUEING NETWORKS USING MAX WEIGHT ALGORITHM M.Sathish Kumar *, G.Sathish Kumar * Department
More informationOnline Algorithms for Energy Cost Minimization in Cellular Networks
Online Algorithms for Energy Cost Minimization in Cellular Networks Ali Abbasi and Majid Ghaderi Department of Computer Science, University of Calgary {aabbasi, mghaderi}@ucalgary.ca Abstract Dynamic base
More informationOptimal Risk-aware Power Procurement for Data Centers in Day-Ahead and Real-Time Electricity Markets
Optimal Risk-aware Power Procurement for Data Centers in Day-Ahead and Real-Time Electricity Markets Mahdi Ghamkhari, Hamed Mohsenian-Rad, and Adam Wierman Department of Electrical Engineering, University
More informationGrid Connected Energy Storage for Residential, Commercial & Industrial Use - An Australian Perspective
IEA Storage Workshop February 2013 Grid Connected Energy Storage for Residential, Commercial & Industrial Use - An Australian Perspective Tony Vassallo Faculty of Engineering & Information Technologies
More informationOPTIMAl PREMIUM CONTROl IN A NON-liFE INSURANCE BUSINESS
ONDERZOEKSRAPPORT NR 8904 OPTIMAl PREMIUM CONTROl IN A NON-liFE INSURANCE BUSINESS BY M. VANDEBROEK & J. DHAENE D/1989/2376/5 1 IN A OPTIMAl PREMIUM CONTROl NON-liFE INSURANCE BUSINESS By Martina Vandebroek
More informationCoordination of Cloud Computing and Smart Power Grids
Coordination of Cloud Computing and Smart ower Grids Amir-Hamed Mohsenian-Rad and Alberto Leon-Garcia Department of Electrical and Computer Engineering, University of Toronto, Toronto, Canada e-mails:
More informationOn the Interaction and Competition among Internet Service Providers
On the Interaction and Competition among Internet Service Providers Sam C.M. Lee John C.S. Lui + Abstract The current Internet architecture comprises of different privately owned Internet service providers
More informationGlobal Cost Diversity Aware Dispatch Algorithm for Heterogeneous Data Centers
Global Cost Diversity Aware Dispatch Algorithm for Heterogeneous Data Centers Ananth Narayan S. ans6@sfu.ca Soubhra Sharangi ssa121@sfu.ca Simon Fraser University Burnaby, Canada Alexandra Fedorova fedorova@cs.sfu.ca
More informationExploring Smart Grid and Data Center Interactions for Electric Power Load Balancing
Exploring Smart Grid and Data Center Interactions for Electric Power Load Balancing Hao Wang, Jianwei Huang, Xiaojun Lin, Hamed Mohsenian-Rad Department of Information Engineering, The Chinese University
More informationFalloc: Fair Network Bandwidth Allocation in IaaS Datacenters via a Bargaining Game Approach
Falloc: Fair Network Bandwidth Allocation in IaaS Datacenters via a Bargaining Game Approach Fangming Liu 1,2 In collaboration with Jian Guo 1,2, Haowen Tang 1,2, Yingnan Lian 1,2, Hai Jin 2 and John C.S.
More informationDual-side Dynamic Controls for Cost Minimization in Mobile Cloud Computing Systems
Dual-side Dynamic Controls for Cost Minimization in Mobile Cloud Computing Systems Yeongin Kim, Jeongho Kwak and Song Chong Department of Electrical Engineering, KAIST E-mail: y.kim, h.kwak}@netsys.kaist.ac.kr,
More informationStability Analysis of Intelligent Home Energy Management System(HEMS)
Stability Analysis of Intelligent Home Energy Management System(HEMS) Saher Umer s1120201 saher@jaist.ac.jp 2010 Cisco Systems, Inc. All rights reserved. 1 Presentation Outline 1 1- Introduction 1.1- Home
More informationOnline Dynamic Capacity Provisioning in Data Centers
Online Dynamic Capacity Provisioning in Data Centers Minghong Lin and Adam Wierman California Institute of Technology Lachlan L. H. Andrew Swinburne University of Technology Eno Thereska Microsoft Research
More informationOn the Cost-QoE Trade-off for Cloud Media Streaming under Amazon EC2 Pricing Models
On the Cost-QoE Trade-off for Cloud Media Streaming under Amazon EC2 Pricing Models Jian He, Yonggang Wen, Jianwei Huang,DiWu Department of Computer Science, Sun Yat-Sen University, China School of Computer
More informationOptimal Online Multi-Instance Acquisition in IaaS Clouds
Optimal Online Multi-Instance Acquisition in IaaS Clouds Wei Wang, Student Member, IEEE, Ben Liang, Senior Member, IEEE, and Baochun Li, Fellow, IEEE Abstract Infrastructure-as-a-Service (IaaS) clouds
More informationOptimal Power Flow Analysis of Energy Storage for Congestion Relief, Emissions Reduction, and Cost Savings
1 Optimal Power Flow Analysis of Energy Storage for Congestion Relief, Emissions Reduction, and Cost Savings Zhouxing Hu, Student Member, IEEE, and Ward T. Jewell, Fellow, IEEE Abstract AC optimal power
More informationOptimal Allocation of renewable Energy Parks: A Two Stage Optimization Model. Mohammad Atef, Carmen Gervet German University in Cairo, EGYPT
Optimal Allocation of renewable Energy Parks: A Two Stage Optimization Model Mohammad Atef, Carmen Gervet German University in Cairo, EGYPT JFPC 2012 1 Overview Egypt & Renewable Energy Prospects Case
More informationEnabling Datacenter Servers to Scale Out Economically and Sustainably
Enabling Datacenter Servers to Scale Out Economically and Sustainably Chao Li, Yang Hu, Ruijin Zhou, Ming Liu, Longjun Liu, Jingling Yuan, Tao Li Chao Li MICRO-46 IDEAL (Intelligent Design of Efficient
More informationA MANAGER S ROADMAP GUIDE FOR LATERAL TRANS-SHIPMENT IN SUPPLY CHAIN INVENTORY MANAGEMENT
A MANAGER S ROADMAP GUIDE FOR LATERAL TRANS-SHIPMENT IN SUPPLY CHAIN INVENTORY MANAGEMENT By implementing the proposed five decision rules for lateral trans-shipment decision support, professional inventory
More informationData centers & energy: Did we get it backwards? Adam Wierman, Caltech
Data centers & energy: Did we get it backwards? Adam Wierman, Caltech The typical story about energy & data centers: The typical story about energy & data centers: Sustainable data centers Remember: The
More informationKeywords: Dynamic Load Balancing, Process Migration, Load Indices, Threshold Level, Response Time, Process Age.
Volume 3, Issue 10, October 2013 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com Load Measurement
More informationA Hierarchical Structure based Coverage Repair in Wireless Sensor Networks
A Hierarchical Structure based Coverage Repair in Wireless Sensor Networks Jie Wu Computer Science & Engineering Department Florida Atlantic University Boca Raton, FL 3343, USA E-mail: jie@cse.fau.edu
More informationGreenColo: Incentivizing Tenants for Reducing Carbon Footprint in Colocation Data Centers
1 GreenColo: Incentivizing Tenants for Reducing Carbon Footprint in Colocation Data Centers Mohammad A. Islam, Hasan Mahmud, Shaolei Ren, Xiaorui Wang, Haven Wang, Joseph Scott Abstract The massive energy
More informationInternational Journal of Advanced Research in Computer Science and Software Engineering
Volume 2, Issue 9, September 2012 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com An Experimental
More informationRELIABILITY OF ELECTRIC POWER GENERATION IN POWER SYSTEMS WITH THERMAL AND WIND POWER PLANTS
Oil Shale, 27, Vol. 24, No. 2 Special ISSN 28-189X pp. 197 28 27 Estonian Academy Publishers RELIABILITY OF ELECTRIC POWER GENERATION IN POWER SYSTEMS WITH THERMAL AND WIND POWER PLANTS M. VALDMA, M. KEEL,
More informationMulti-Faceted Solution for Managing Flexibility with High Penetration of Renewable Resources
Multi-Faceted Solution for Managing Flexibility with High Penetration of Renewable Resources FERC Technical Conference Increasing RT & DA Market Efficiency Through Improved Software June 24 26, 2013 Nivad
More informationBig Data Collection and Utilization for Operational Support of Smarter Social Infrastructure
Hitachi Review Vol. 63 (2014), No. 1 18 Big Data Collection and Utilization for Operational Support of Smarter Social Infrastructure Kazuaki Iwamura Hideki Tonooka Yoshihiro Mizuno Yuichi Mashita OVERVIEW:
More informationFunctional Optimization Models for Active Queue Management
Functional Optimization Models for Active Queue Management Yixin Chen Department of Computer Science and Engineering Washington University in St Louis 1 Brookings Drive St Louis, MO 63130, USA chen@cse.wustl.edu
More informationEnergy Benefit of Network Coding for Multiple Unicast in Wireless Networks
Energy Benefit of Network Coding for Multiple Unicast in Wireless Networks Jasper Goseling IRCTR/CWPC, WMC Group Delft University of Technology The Netherlands j.goseling@tudelft.nl Abstract Jos H. Weber
More informationPeer-Assisted Online Storage and Distribution: Modeling and Server Strategies
Peer-Assisted Online Storage and Distribution: Modeling and Server Strategies Ye Sun, Fangming Liu, Bo Li Hong Kong University of Science & Technology {yesun, lfxad, bli}@cse.ust.hk Baochun Li University
More informationBig Data Processing of Data Services in Geo Distributed Data Centers Using Cost Minimization Implementation
Big Data Processing of Data Services in Geo Distributed Data Centers Using Cost Minimization Implementation A. Dhineshkumar, M.Sakthivel Final Year MCA Student, VelTech HighTech Engineering College, Chennai,
More informationEnergy harvesting Communication networks: OPtimization and demonstration E-CROPS
Energy harvesting Communication networks: OPtimization and demonstration presented by Imperial College London March 27, 2013 Motivation Project title: Energy harvesting Communication networks: OPtimization
More informationENERGY SAVING SYSTEM FOR ANDROID SMARTPHONE APPLICATION DEVELOPMENT
ENERGY SAVING SYSTEM FOR ANDROID SMARTPHONE APPLICATION DEVELOPMENT Dipika K. Nimbokar 1, Ranjit M. Shende 2 1 B.E.,IT,J.D.I.E.T.,Yavatmal,Maharashtra,India,dipika23nimbokar@gmail.com 2 Assistant Prof,
More informationDesigning and Managing Datacenters Powered by Renewable Energy
Designing and anaging Datacenters Powered by Renewable Energy Íñigo Goiri, William Katsak, Kien Le, Thu D. Nguyen, and Ricardo Bianchini Department of Computer Science, Rutgers University {goiri,wkatsak,lekien,tdnguyen,ricardob}@cs.rutgers.edu
More informationADAPTIVE LOAD BALANCING ALGORITHM USING MODIFIED RESOURCE ALLOCATION STRATEGIES ON INFRASTRUCTURE AS A SERVICE CLOUD SYSTEMS
ADAPTIVE LOAD BALANCING ALGORITHM USING MODIFIED RESOURCE ALLOCATION STRATEGIES ON INFRASTRUCTURE AS A SERVICE CLOUD SYSTEMS Lavanya M., Sahana V., Swathi Rekha K. and Vaithiyanathan V. School of Computing,
More informationCURTAIL THE EXPENDITURE OF BIG DATA PROCESSING USING MIXED INTEGER NON-LINEAR PROGRAMMING
Journal homepage: http://www.journalijar.com INTERNATIONAL JOURNAL OF ADVANCED RESEARCH RESEARCH ARTICLE CURTAIL THE EXPENDITURE OF BIG DATA PROCESSING USING MIXED INTEGER NON-LINEAR PROGRAMMING R.Kohila
More informationSPARE PARTS INVENTORY SYSTEMS UNDER AN INCREASING FAILURE RATE DEMAND INTERVAL DISTRIBUTION
SPARE PARS INVENORY SYSEMS UNDER AN INCREASING FAILURE RAE DEMAND INERVAL DISRIBUION Safa Saidane 1, M. Zied Babai 2, M. Salah Aguir 3, Ouajdi Korbaa 4 1 National School of Computer Sciences (unisia),
More informationVOLATILITY AND DEVIATION OF DISTRIBUTED SOLAR
VOLATILITY AND DEVIATION OF DISTRIBUTED SOLAR Andrew Goldstein Yale University 68 High Street New Haven, CT 06511 andrew.goldstein@yale.edu Alexander Thornton Shawn Kerrigan Locus Energy 657 Mission St.
More informationDynamic Workload Management in Heterogeneous Cloud Computing Environments
Dynamic Workload Management in Heterogeneous Cloud Computing Environments Qi Zhang and Raouf Boutaba University of Waterloo IEEE/IFIP Network Operations and Management Symposium Krakow, Poland May 7, 2014
More informationEnergy-Aware Multi-agent Server Consolidation in Federated Clouds
Energy-Aware Multi-agent Server Consolidation in Federated Clouds Alessandro Ferreira Leite 1 and Alba Cristina Magalhaes Alves de Melo 1 Department of Computer Science University of Brasilia, Brasilia,
More informationCOST MINIMIZATION OF RUNNING MAPREDUCE ACROSS GEOGRAPHICALLY DISTRIBUTED DATA CENTERS
COST MINIMIZATION OF RUNNING MAPREDUCE ACROSS GEOGRAPHICALLY DISTRIBUTED DATA CENTERS Ms. T. Cowsalya PG Scholar, SVS College of Engineering, Coimbatore, Tamilnadu, India Dr. S. Senthamarai Kannan Assistant
More information