network at the access level by using solar energy which is renewable and clean. Keywords- LR-PON, telecom network, access network, solar energy, admission ...
Building Energy Efficient LR-PON For Desert Terrain of Saudi Arabia Mohammad Siraj1, Amanullah Fatehmulla2, Saleh Alshebeili3, Kamalrulnizam Abu Bakar4 1,4
Faculty of Computer Science and Information Systems , University Technologi Malaysia Department of Physics, College of Science P.O. Box 2455,King Saud University, Riyadh, Saudi Arabia 1,3 PSATRI/STC-Chair, College of Engineering, King Saud University, P.O. Box 800, Riyadh, Saudi Arabia 2
Abstract— With the increase of bandwidth demands, an increasing concern prevails among researchers towards energy consumption in next generation access networks. Attempts have been made by researchers with an overview of techniques and architectures for energy consumption minimization in core networks. Growing concern equally exists essentially to reduce Green House Gases (GHG) emissions from computers and network equipments. In the present paper, we focus on renewable energy utilization at access level as well with a view to build energy efficient long range passive optical networks (LRPON) for desert terrain of Saudi Arabia. In this paper we have presented LR-PON model between Riyadh and Al-Kharj with an admission control algorithm for efficient performance of LRPON. Also, in this paper we have shown how to reduce network complexities and provide a cost effective solution by greening the network at the access level by using solar energy which is renewable and clean. Keywords- LR-PON, telecom network, access network, solar energy, admission control
I.
INTRODUCTION
Energy conservation has become one of the most sought topics of the present era. Research directions are being explored to address this situation. This motivation is driven mainly because of two reasons. The primary reason is that hydrocarbon energy, which is not renewable, ultimately will be consumed in the distant future. The second reason is to reduce the carbon footprint. Information and Communication Technology (ICT), which is a rapidly growing technology, is one of the major consumers of energy. In fact, as per the data of 2009, ICT consumes about 8% of the total electricity consumption of the world [1]. In this paper we focus on energy efficiency of access network which is the last part of telecommunication networks. The access network connects the telecom Central Office to the business and residential consumers. Our focus is on access networks, as the access network consumes around 70 % of telecommunication network energy consumption. This number will grow in the coming years with the rapid increase in number of subscribers and the broadband network demand [2]. This consumption in access network is due to the presence of large number of active elements [3]. Due to the continuous growth of bandwidth requirement, the energy consumption of computers and network devices have risen [4], [5], [6]. Consciousness on energy consumption is rising nowadays in the Information and Communication Technology. In the present era energy
consumption have become an environmental issue and therefore social and political issue. So to make our planet Green, carbon footprint has to be reduced. We have to tackle this issue by designing efficient energy schemes for future telecommunication networks. Through these efficient energy schemes in future access networks, the energy cost will come not only down but also we will be moving towards making our planet green. There are various access technologies available in the market. DSL (Digital Subscriber Line), FTTH (Fiber to the Home), Wireless and Cellular Networks, WOBAN (Wireless Optical Broadband Access Networks) are the present day technologies. Currently, in access networks, there is a trend towards FTTH because of requirement of high data rates. FTTH are being preferred because they have low attenuation and provide higher bandwidth. The higher data rate demand is coming from applications such as video on demand, HDTV, online games etc. After the deployment of Fiber in FTTH access networks the network is simplified and the total cost of ownership (TOC) comes down [7]. Thus consolidation of the networks cuts the energy costs. It is widely seen that electricity is lost considerably in access network equipments. This is due to various reasons. Firstly poor network design is a cause of electricity loss. Secondly idle elements [8] in the network consume significant energy. It is estimated that networking equipments are utilized not more than 15% [9]. So there is a need that network is utilized properly to reduce the energy consumption of these idle elements. The energy consumption of the network components such as routers, switches, Optical Cross-Connect, optoelectronic Switch etc. is quite high. For e.g. Cisco CRS-1 Core Router whose capacity is 92 Tbps consumes 1020 kW [10].Alcatel-Lucent 1675 Optoelectronic Switch whose capacity is 1.2 Tbps consumes 2.5 kW [11]. Researchers are working on two directions to reduce the consumption of energy in these idle elements. The two approaches are: a) energy saving protocol design [12],b) energy saving network design. In this paper we have proposed an architecture for deploying LR-PON in the desert terrain of Saudi Arabia between Riyadh and Al Kharj and the distance between them 90 kilometers. Also in this paper we have proposed solar energy to power our network elements of access networks which is clean and renewable. We have presented an admission control algorithm for LR-PON for efficient performance.
The paper is organized as follows. The following section discusses related work. Section III describes about the Network Structure. In section IV and V we describe about our Network Model and Energy Model. Section VI discusses Analytical Model .Finally section VII concludes this work. II.
RELATED WORK
In recent years researchers have published works related to energy consumption of telecom networks and how to improve energy efficiency in light of growing broadband demand and increase usage of internet. For example, the power consumption distribution in networks has been estimated [13], as well as the energy consumption of the internet [14]. An overall energy forecast for Telecommunication network during the operation phase when delivering broadband services delivery has been done [15] Wireless networking community have designed energy efficient protocols [16]. They have done the comparison between different emerging technologies such as passive optical network (PON), fiber to the node (FTTN), Point to Point Optical network (PtP) and WiMAX. There have been several proposal and recommendation for energy efficient PON’s. There have been proposal for low power (sleep) for
PON equipments [17]. Handshaking protocol for coordinated sleeping mechanism [18], shedding power in user network interface (UNI) and access network interface (ANI) [19] etc. Hybrid Wireless Optical Broadband Access Network (WOBAN) [20] which is a combination of Passive Optical Network (PON) and WiMAX has also been proposed towards energy efficient design. Some researchers have focused energy efficiency in access networks [3] [21].Once these suggestions are implemented energy can be saved. Despite all these efforts, there remains a significant challenge to deploy energy efficient design in access networks. III.
NETWORK STRUCTURE
In recent years researchers have published works related to energy consumption of telecom networks and how to improve efficiency in telecom network. A Telecom Network can logically be divided into three main domains-the core network, the metropolitan network and access network as shown in fig.1 below
Figure1. Schematic Diagram of a Telecom Network with different access network options In this section we describe about these three domains and discuss essential network elements. Core Network: The term core network is referred to the backbone of the telecom network. They connect cities and spans continental distances. The core network comprises of large routers to perform all the necessary routing. Generally the core routers are connected in mesh architecture. In the core network, optical technologies are widely used to support the physical infrastructure to achieve high speed, high capacity etc. High Capacity Wavelength Division Multiplexing (WDM) fiber links are used to interconnect these routers and to connect to other network operators. Network architectures based on IP over SONET/SDH over WDM or IP over WDM have been deployed. Typically WDM links in today’s networks comprises of 10 Gb/s Packet over SONET (PoS), or 10 Gb Ethernet, or 40 Gb/s PoS.
Metropolitan Network: The metro network serves as an interface between core and the access network. Typically they cover metropolitan region. The essential components are Ethernet switches, broadband network gateways, edge routers etc. The edge switch typically connects an end user's LAN to an ISP's network or backbone. Sometimes an edge switch is also called an access node or a service node. Generally their capacity is from 2.5 to 480 Gbps. An edge router is sometimes referred to as a boundary router. They are designed to support high traffic volumes. Broadband network gateways perform IP routing, secure admission control, protection from malicious attacks (DDOS), secure Internet tunneling and connect to other edge routers. The edge switches and routers are designed to improve performance by providing scalability, reliability and security for small and large networks. OADMs (Optical Add Drop Multiplexer), SONET
symmetric transmission rate for upstream and downstream directions. The core exchange would be placed in Riyadh and local exchange at Al Kharj. A single fiber would run from the local exchange on the access network to a location close to the customer premises. It is shared to 1024 users by a 1024 split. This 1024 split is made up of a cascade of two N: 16 and one N: 4 splits [23]. The distribution section is divided into three parts each containing a length of fibre (3km, 3km and 4km) making 10 km. Though increasing the data rate to 10 Gbps will need an expensive high speed transmitter for each ONU but the cost is compensated as the OLT is shared by 1024 users. The reason why we have chosen LR-PON is because a) it has passive elements b) of its reach many remote locations can be combined, thereby reducing energy consumption in access network.
ADM (Add Drop Multiplexer) form part of Metro Optical Network. IV.
NETWORK MODEL
In this section we discuss our model of the access network: Long reach Passive Optical Network (LR- PON): LR-PON is a novel technology aimed at replacing the current copper access network with a shared fibre network offering much greater capacity and is proposed as a cost effective solution for future broadband optical networks. LR-PON use Optical amplifiers (OA). LR-PON single segment coverage area is 100 km. It can be extended beyond 100km by utilizing Wavelength dense multiplexing (WDM) technology. In our model fig.2, we consider 1024 way split 100 km LR PON with 10 Gbps
. Figure 2. Proposed Architecture for Long Reach Passive Optical Networks V.
(Green) system using PV modules (36 modules each of 220 W) can be designed. In this paper, an energy efficient LR-PON model at the access level for desert terrain habitants of Saudi Arabia has been developed keeping in view of the expected number of subscribers and their energy consumption towards the network components (Table I)
ENERGY EFFICIENT LR-PON MODEL USING SOLAR ENERGY:
Stand-alone PV(Photovoltaic) systems have shown to be reliable and cost effective for various applications and have attracted the users. For a specific application with an estimated requirement of energy consumption, a renewable energy
Table I ENERGY CONSUMPTION PER CUSTOMER
TOTAL ENERGY CONSUMPTION TECHNOLOGY
PTU(kW)
NTU
PRN(W)
NRN
PCPE(W)
Technology Limit
Maximum AT
Power Per Customer
Min. Energy Per Bit
ADSL
1.73[11]
1024
0
N/A
5
15 Mb/s
2 Mb/s
7.8 W
3.8µJ
PtP
4.72[25]
1024
0
N/A
4[26]
125 Mb/s
12.2 W
0.1 µJ
PON
1.24[27]
1024
0
N/A
5[28]
2.4 Gb/s
16 Mb/s
7.6 W
0.5 µJ
FTTN
1.34[27]
1024
47[29]
16[30]
10[31]
50 Mb/s
2 Mb/s
13.2 W
6.6 µJ
1 Gb/s
Pa = PCPE +
PRN 2 PTU + N RN NTU
(1)
where PCPE, PRN, and PTU are the powers consumed by the customer premises equipment (i.e., the modem), the remote node and the terminal unit at the central office, respectively. NRN and NTU are the number of customers or subscribers that share a remote node and the number of customers that share a terminal unit, respectively. The last term on the right hand side of equation (1) includes a factor of 2 to account for additional overheads such as external power supplies and cooling requirements [24]. Table I lists the values of the parameters for each access technology. On the left-hand side of this table are the parameters referred to in this section and on the right-hand side the parameters are normalized on a per customer basis. The number of customers per Cisco 4503 Ethernet switch NTU(PtP) is given by
(116 − LB ) Gb / s ( 64 − LB ) Gb / s NTU ( PtP ) = min ' 1Gb / s + AT 2 AT
The energy cost in US $ --------$6117.98 /year In view of the above calculations and the initial cost of our PV system including the initial electrical installation cost power the said LR-PON, a plot showing the comparison of the energy cost incurred with PV and conventional systems and years of operation has been drawn(Fig.3 ). Electrical Energy Cost Cost of Solar Energy
120000
100000
80000
Cost in US $
The per customer power consumption of all four access technologies Pa can be expressed as
60000
40000
payback period
20000
0
(2) where the first term is due to the limit of port capacity, and the second term is due to the limit of switching capacity.
Total PTU= ADSLPTU + PtPPTU + PONPTU + FTTNPTU = (1.73 + 4.72
+ 1.24 + 0.17) kW
= 7.76 kW . Based on the above energy consumption, solar energy (Renewable energy source) model has been designed to power the LR-PON at access level. Cost comparison between electrical & solar energy: (Consumption at Access Network) Total Power needed at Access level = 7.76 kW Calculation of cost of solar energy needed to power LR-PON: About 36 solar modules each of 220W are enough which cost about US$ 45000 (cost of PV system+ complete solar panel grid-tie systems with battery backups) The energy needed daily ------ 7.76 kWh × 18 hours = 139.68 kWh/day (Note: 5- 6 hours/day excluded in charging process) The average cost of electrical energy ≈ 12cents/kWh (Approximate value in KSA) So, the daily energy cost ------- 139.68 kWh/day × 12cents/ Kwh = 1676.16 cents/day The energy cost per year -------1676.16 cents/day × 365 days/year ≈ 611,798.4cents/year
0
2
4
6
8
10
12
14
16
Years of Operation
Figure 3. Cost comparison between the PV and Conventional energy in operating LR-PON system From figure 3, it is clear that the Solar Energy and electrical energy costs intersect at about 5½ years of operation which is the payback period. Since the maintenance cost is low or approximately nil, it appears that the solar energy becomes almost free after 5½ years and it demonstrates the economic effectiveness. Therefore, electrical energy will cost more than solar energy, especially for large scale application the cost will increase rapidly. In addition to that the solar energy is clean and safe. Our experience of the design and development of PV low power water pumping system with relatively low cost suggests that this small scale technology can serve the purposes like lifting water to 10 m head, small area irrigation especially in remote rural areas and in deserts[32]. Also it supports our experience that by the use of renewable energies in windy and sunny areas a simple, cheap and reliable irrigation (pumping) or cooling (refrigerating) systems can be applied, avoiding the disadvantages of the already existed diesel-powered systems [33]. However, photovoltaic powered Network systems have reached a technical maturity and hence the design of power consumption model at access networks using solar energy becomes a viable option and is characterized of a high reliability. Further, it offers clean energy free from carbon footprint (GHG emissions) and provides the advantages of reduced operation and maintenance cost. VI.
ANALYTICAL MODEL
Analytical Model of Admission Control:
18
20
The proposed model imposes minimum additional burden on the server and the network. Let the maximum number of requests that can be serviced at the server at Central Office (CO) be n. The central office (CO) connects the core network and the access network, and implements layer 2 and layer 3 functions, e.g., resource allocation, service aggregation, management, and control. Let Pb be the overall bandwidth and Pr is the request from the PON segments. If
n≤
Pb Pr
pa∗+1 . If this condition is not met, then request Q +2 will be i checked and so on. If the request can’t be admitted by all the sections, then the request will be considered blocked and the server will discard the request. The flowchart for the admission control algorithm is shown below in fig.5
(3)
It means the requests can be handled by the server buffer otherwise there is congestion. So we can say this is the admission control test at the CO server that the server is capable of serving maximum n requests. For efficient performance we have assumed the servers and clients are distributive. Let the buffer capacity at CO be C .Let us have k service requests from different distribution segment with different data rate λi where i=1 to k. The server capacity is divided into many sections with each section port having capacity Cj where j=1 to k as shown in figure below. Total system capacity is
Figure 5. Admission Control Flowchart VII. CONCLUSION AND FUTURE WORK
k
C = ∑Cj
(4)
j =1
This paper shows the feasibility of reducing carbon foot print in access networks through the use of solar energy.The calculations of the specific energy consumption costs show the promising economic effectiveness and reliability of the designed solar energy conversion system for LR-PON. A key parameter that will effect growth of energy consumption of the network is the rate at which the energy efficiency of the network equipment in the infrastructure improves with successive generations of new technology. The overall energy efficiency of the network will depend on how rapidly network operators choose to replace older equipment with newer technology. In future work, if any impairment arises from bursty traffic it needs to be addressed. REFERENCES
Figure 4.Traffic Arrival
[1]
Request from the distribution segment x will be admitted to a partition y in the server with probability k
p ( x, y ) = p Where *
*
∑p i =1
∗ i
[2]
=1 (5)
A new request with Poisson Distribution will come from the distribution segment. The streaming Flow will be entertained as long as it satisfies eq.1. We assume the server ports occupancy for request i is Ci. When a new request i arrives, we check whether Qi < Ci, if so, then it is admitted with probability P*x. If request is served, then Qi = Qi+1. We then check Qi+1 < Ci+1. This request will be admitted in the server with probability
[3] [4]
[5]
P. Leisching and M. Pickavet, “Energy footprint of ICT: Forecasts and network solutions,” OFC/NFOEC’09, Workshop on Energy Footprint of ICT: Forecast and Network Solutions, San Diego, CA, Mar. 2009. C. Lange and A. Gladisch, “Energy consumption of telecommunication networks - a network operator’s view,” OFC/NFOEC’09, Workshop onEnergy Footprint of ICT: Forecast and Network Solutions, San Diego, CA, March 2009. C. Lange and A. Gladisch, “On the energy consumption of FTTH access networks,” presented at the OFC/NFOEC, San Diego, CA, Mar.2009. W. Vereecken, L. Deboosere, D. Colle, B. Vermeulen, M. Pickavet, B.Dhoedt, and P. Demeester, “Energy efficiency in telecommunication networks,” European Conference on Networks and Optical Communications & Optical Cabling and Infrastructure (NOC’08), July 2008. M. Pickavet, W. Vereecken, S. Demeyer, P. Audenaert, D. Colle, C.Develder, and P. Demeester, “Contribution and role of network
[6]
[7]
[8]
[9]
[10]
[11] [12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23] [24]
architectures in the footprint reduction of ICT,” NOC/OC&I’09, Valladolid, Spain, June 2009. M. Pickavet, W. Vereecken, S. Demeyer, P. Audenaert, B. Vermeulen, C. Develder, D. Colle, B. Dhoedt, and P. Demeester, “Worldwide energy needs for lCT: the rise of power-aware networking,” International Symposium on Advanced Networks and Telecommunication Systems (ANTS’08), Bombay, India, Dec. 2008. C. Lange,R.H¨ulsermann, D.Kosiankowski, F.Geilhardt, and A. Gladisch,“Effects of network node consolidation in optical access and aggregation networks on costs and power consumption,” presented at the SPIE Photonics West—Optical Communications: Systems and Subsystems, San Francisco, CA, Jan. 23–28, 2010. K. J. Christensen, C. Gunaratne, B. Nordman, and A. D. George, “The next frontier for communications networks: Power management,” Elsevier Comput. Commun. vol. 27, pp. 1758–1770, Aug. 2004. J. Russo, “Network technology energy efficiency,” presented at the Berkeley Symp. Energy Efficient Electron. Syst., Berkeley, CA, Jun.2009. R. S. Tucker, “Energy footprint of the network,” OFC/NFOEC’09, Workshop on Energy Footprint of ICT: Forecast and Network Solutions, San Diego, CA, Mar. 2009 “Alcatel-Lucent1675 Lambda Unite MSS datasheet,”http://www.alcatelLucent.com/wps/ J. Chabarek, J. Sommers, P. Barford, C. Estan, D. Tsiang, and S.Wright, “Power awareness in network design and routing,” in Proc.IEEE INFOCOM, Phoenix, AZ, 2008, pp. 1130–1138. M. Pickavet, W. Vereecken, S. Demeyer, P. Audenaert, B. Vermeulen,C. Develder, D. Colle, B. Dhoedt, and P. Demeester, “Worldwide energy needs for ICT: The rise of power-aware networking,” presented at the Int. Symp. Advanced Networks and Telecommunication Systems (ANTS), Bombay, India, Dec. 15–17, 2008. J. Baliga, R. Ayre, K. Hinton, W. V. Sorin, and R. S. Tucker, “Energy consumption in optical IP networks,” J. Lightw. Technol., vol. 27, no. 13, pp. 2391–2403, Jul. 2009. C. Lange, D.Kosiankowski, Rainer Weidmann and Andreas Gladish,”Energy Consumtion of telecommunication Networks and Related improvement Options”IEEE journal of selected topics in quantum Electronics” accepted and to be published in future journal. C. Jones, K. Sivalingam, P. Agrawal, and J. Chen, “A survey of energy efficient network protocols for wireless networks, “Wireless Networks, vol.7, no. 4, pp. 343–358, 2001. J. Mandin, “EPON power saving via sleep mode,” presented at the IEEE P802.3av 10GEPON Task Force Meeting, Sep. 2008 [Online]. available:www.ieee802.org/3/av/public/2008-09/3av-0809-mandin-4.pdf H. Yuanling and M. Hajduczenia, “Adjustable timer value for power saving,” presented at the IEEE P802.3av 10GEPON Task Force Meeting, Sep. 2008 [Online]. Available: www.ieee802.org/3/av/ public/2008-09/3av_0809_hajduczenia_power_Saving.pdf F. J. Effenberger, Opportunities for Power Savings in Optical Access. (2008, Feb.) [Online]. Available: http://www.itu.int/dms-pub/itut/oth/09/05/T09050000010006PDFE.pdf S. Sarkar, S. Dixit, and B. Mukherjee, “Hybrid Wireless-Optical Broadband Access Network (WOBAN): A review of relevant challenges,” Journal of Lightwave Technology (JLT), vol. 25, no. 11, pp. 3329–3340, Nov. 2007 J. Baliga, R. Ayre, W. V. Sorin, K. Hinton, and R. S. Tucker, “Energy consumption in access networks,” presented at the OFC/NFOEC, San Diego, CA, Feb. 2008. P. Chanclou, S. Gosselin, J. F. Palacios, V. L. Alvarez, and E.Zouganeli, “Overview of the optical broadband access evolution: A joint article by operators in the 1ST network of excellence E-photon/one,” IEEE Commun. Mag., vol. 44, no. 8, pp. 29–35, Aug.2006. [23] D. P. Shea and J. E. Mitchell, “A 10 Gb/s 1024-Way Split 100-km Long Reach Optical Access Network,” J. Lightwave Technol., vol. 25, no. 3, pp. 685-693, Mar. 2007
[25] J. Koomey, H. Chong, W. Loh, B. Nordman, and M. Blazek, “Network electricity use associated with wireless personal digital assistants,” ASCE J. Infrastructure Syst., vol. 10, pp. 131–137, Aug. 2004. [26] Cisco Data Sheets, [Online]. Available: http://www.cisco.com [27] Hitachi Data Sheets, [Online]. Available: http://www.hitachi.com [28] Wave7 Data Sheets, [Online]. Available: http://www.wave7optics.com [29] NEC Australia Data Sheets, [Online]. Available: http://www.nec. com.au [30] NEC Japan Data Sheets, [Online]. Available: http://www.nec.co.jp [31] TC Commun. Data Sheets, [Online]. Available: http://www.tccomm.com [32] NEC Australia Oct. 15, 2007, private communication. [33] F.M. Amanullah, A.S. Al-Shammeri and A.M. Al-Dhafiri, “Design, Fabrication and Performance Analysis of Low Power Solar Water Pumping System” Proc. The 8th Arab International Solar Energy Conference & Regional World Renewable Energy Congress Conference & Exhibition, 8-10 March 2004, pp.228-237, Kingdom of Bahrain. [34] Stratis Tapanlis, Lars Beuerman, Tom Scheer and Dimitris Koniarellis, Proceedings of 16th European Photovoltaic Solar Energy Conference, Glasgow, UK, May 1-5 2000, Vol.III, pp 3071-3073
