# A novel parametric analysis of the performance dynamicity of optical network.

INTRODUCTIONThere is a limitation in the Bandwidth allocation till 1980s due to the usage of electrical transmission. The use of optical fiber and optical technologies in the network instead of copper cable immensely increases the transmission speed by terabits per seconds, since light is used by the optical fiber to transmit the data. At the same time, the Bandwidth limitation by the copper is also eliminated by the optical network. In the last few years, due to the grown interest in the field of optical technologies such as switches and routers, advancement in the field of IP networks as well as optical communication takes place. Due to the rapid explosion in the end user, WDM technologies provide the proper platform to transmit high speed multiple data over the single channel [1] [21]. After the introduction of the WDM networks, Optical Burst Switching (OBS) networks, Optical Packet Switching (OPS) and Optical Circuit Switching (OCS) were predominantly used in the optical domain to transmit the data. Since optical packet switching and optical burst switching can travel freewill without static path, Optical packet switching and optical burst switching were highly explored to improve the optical network performance [1]. Due to the exponential growth in the network traffic, there is a need to provide guaranteed QoS in the multifaceted environment. Though various new technologies have been introduced and established in the optical domain, it is mandatory that the network's Quality of Services should be maintained above the threshold value [27]. Quality of Service refers to the overall performance of the network from the user perspective. The Quality of Service can be monitored and maintained by various parameters such as Loss, Delay, and Bandwidth etc. Since today's traffic is dynamic and diverse, these QoS parameters have to be improved to stabilize the network performance and if it is not monitored, the network performance will be unpredicted and hence the network performance will not be maintained to be optimized.

The QoS parameters have been analysed and out of which Hop Count, Wavelength Assignment, Bandwidth and Throughput are the prime essential parameters to be considered in view of the fact that these parameters will impact on the various other parameters and changes the behaviour of the network. More significantly these parameters are interlinked with one another, which mean the characteristic change of one parameter will change the behaviour of the other, thereby will change the overall system performance.

2. Hop Count:

Hop Count refers to the total number of Intermediate devices through which the data travels from source to destination. With respect to the virtual topology, it refers to total number of virtual path found on the route of the virtual path length (or) total number of edges in the respective (or) data driver path on the virtual graph. In the case of designing the network architecture, it is important that there should be a minimal Hop Count for the data to travel from source to destination. Minimizing the total Hop Count will minimize the total intermediate node, so that the data will move with least disturbance from source to destination and with minimal distance [5] [6].

Hop Count will be considered as a minimization objective function (or) constraint in designing the network. We can obtain minimal Total Hop Count parameter ([H.sub.tot] (G, R)) as summation of path exist between source and destination which is proposed in [2]

[H.sub.tot] (G, R) = [summation] [d.sub.G] ([v.sub.i], [v.sub.j])

Where,

[d.sub.G] = distance, [v.sub.i] = source node and [v.sub.j] = destination node.

Though most of the work related to Hop Count as minimization of the distance from the source node to destination, when the data traverse it will be expected to face traffic. It is also consider in the formulation of Hop Count as

[H.sub.tot] (G, R) = [summation] [d.sub.G] ([v.sub.i], [v.sub.j]) [R.sub.ij],

Where, [R.sub.ij] = traffic from source node to destination node.

It is seen that the Hop Count as summation of light tree connecting from source to destination.

Instead of keeping it as minimization function, Hop Count can be considered as a maximization function by the parameter desirability, which considers inversely the length of the shortest path from node n [3].

[[eta].sub.nj-] = 1/F(x), f (x) = length of shortest path from node n.

Hop Count is also considered with traffic bound parameter in [4]. Since the route with least traffic and distance will provide maximum Throughput, the parameter Average Weighted Hop Count is given by

AWHT = [summation] [T.sup.sd.sub.ij] * Hsd / [summation] Tsd

Where,

T = traffic between source and destination

H = Hop Count value.

It is seen that in most of the work, Hop Count is consider to be the parameter for selecting minimum distance metrics and few consider Hop Count with traffic for selecting the path to attain maximum efficient value of data transmission. When the Hop Count increases linearly over distance, the Average load or traffic will increases linearly. The variation of average hyper edge load with respect to Hop Count will increase with respect to increase in Hop Count is given below [22]

Dropping probability is also an important criterion which affects the performance of the network. The Dropping probability increases with respect to increase in the value of the Hop Count. The Dropping Probability of Just in Time (JIT), Balanced Just in Time (BJIT) and prioritized random early discard (PRED) on US Long Haul with different loads variation (W=64) increases with respect to Hop Count is given below [23].

3. Wavelength Assignment:

Wavelength Division Multiplexing plays a vital role in building the next generation optical networks. WDM technique allows several wavelengths to be shared and transmitted over the optical fibre. Assigning the wavelength from source to destination node to transmit the signal will be a challenging task. Due to the dynamic nature of the network, the same wavelength cannot be expected to be assigned from source to destination and hence wavelength continuity constraint cannot be satisfied. So several wavelength converters along with Wavelength Assignment techniques are required to meet the dynamic needs [12].

For the wavelength assignment, three commonly used algorithms are First Fit (FF), Least Used (LU) and Most Used (MU). First Fit algorithm number all the available wavelength, so that when the demand occurs, the algorithm assigns the smaller wavelength number to the network. Least Used method assigns the least used wavelength in order to balance the load. Most Used algorithm assigns most used wavelength in the network. The blocking probability and throughput of first fit shows better performance for single fiber and 4 wavelengths, when compared to other algorithms [24].

The various heuristic algorithms have been proposed for assigning the wavelength in the optical networks for the wavelength continuous routing [7].The commonly used two Greedy algorithms are expected to produce efficient output based on the wavelength mesh graph, One is based on the shortest path named as shortest path preference in which the wavelengths are assigned with respect to the shortest path first and other algorithm is based on fixed predetermined sequence in which Wavelength Assignment will follow the predefined order. Similarly, with respect to Non wavelength continuous network, two algorithms are proposed, one is Exhaustic algorithm which uses the shortest path and any available wavelength in the path to assign the wavelength and other algorithm is least loaded algorithm which considers the current traffic followed by the shortest path to assign the wavelength.

For the wavelength continuity network, first fit preference shows superior performance with respect to blocking probability. For the wavelength non-continuity network, least loaded algorithm shows the superior result. Whereas, for considering both the cases, greedy shortest path preference shows the superior result. The various wavelength reassignment algorithms were proposed in which the reassignment scheme is followed to accommodate the blocked call [9]. Least congested overlapped reallocates the overall link from least congested wavelength (LC-NLC) to next least congested wavelength using first fit performs well with respect to blocking probability then LC-NLC-shortest path and Random reassignment scheme. Least congested path will be determined based on link utilization metric (U) which is given by

Label (L) = [summation over (k)] [summation over (l)] [Time slot in use / l * f * k]

Where,

l = number of links = number of time slot per [lambda] and k = number of [lambda] per link.

The Best-fit Wavelength Assignment algorithm allocates adequate wavelength to the primary path by means of minimal cost function [10].The generic objective function is also considered to allocate the wavelength [8].The generic objective function consider the label value to allocates the wavelength ,which is given by

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII],

Where,

[W.sub.[lambda]x] = weight attribute for [lambda]

l = link connecting between 2 nodes

n = nodes.

The hop length can also be related to the Weight Factor such that

Weight Factor = [f.sub.P] / w-[f.sub.P] [square root of ([L.sub.P])]

Where,

[f.sub.P] = no. of free path

w = wavelength

[L.sub.P] = hop length.

Hence it is concluded that most of the articles use first fit, random select, shortest path and least congested criteria in selecting the path and few articles focus on the weight factor (or) objective function in assigning the wavelength. It is also noted that the weight function focuses on the least congestion and the Hop Count as well.

4. Bandwidth:

Bandwidth estimation and Bandwidth allocation is the imperative component to provide Quality of Services in the optical network. The allocation of Bandwidth will indirectly improve the system performance by improves its Throughput and reduces congestion and delay. Normally, Bandwidth allocation will be based on static as well as dynamic scheme. It is proposed that too much of Bandwidth is wasted with respect to static Bandwidth Allocation, whereas Dynamic Bandwidth Allocation (DBA) allocates Bandwidth based on requirement [11] [15].

In the flexible Bandwidth optical network, Bandwidth Blocking Probability should be reduced to achieve high utilization of bandwidth. Various algorithms have been proposed for spectrum allocation and collusion reduction. Fixed Shortest Path with Collusion Aware (FSP-CA) routing algorithm, Fixed Shortest Path with First Fit (FSP-FF), Most Common Available Patterns with Hop Constraint and Collusion Aware routing algorithm (HC-MCAP-CA), Most Common Available Patterns with Hop Constraint and First Fit algorithms (HC-MCAP- FF) will be commonly employed to transmit data.HC-MCAP-CA achieves better performance with in both light load and highly dynamic traffic by efficient load balancing and collusion avoidance [25]

In the existing passive optical network, which influences the existing successful access network, Bandwidth allocation will be a challenging task. N.Moradpoor et al. analysed the various access network and concluded that expedited forwarding along with DBA performs well with respect to average queuing delay compared with other schemes. It is known that the granted Bandwidth as the summation of Minimum Bandwidth and Excess Bandwidth [11].

Where Minimum bandwidth is given by

[B.sub.min (i)] = [[[alpha].sub.i] x (Tcycle - N x Tguard) x R / 8, [summation]] [[alpha].sub.i] = 1

Where,

[T.sub.cycle] is Maximum transmission cycle

T guard is the guard time between two consecutive time slots

N = number of Optical Network Unit

R = transmission Rate (Mb/s).

In order to improve the Bandwidth utilization under low and medium traffic load, probability density function (PDF) of the nth ONU is calculated such that granted Bandwidth is equal to required Bandwidth and PDF (n) is greater than the Threshold.

Hence

PDF [(n).sub.t-1] = granted BW [(n).sub.t-1] / [summation] total Bandwidth allocation in (t -1)

The various algorithms were analysed to improve Throughput in the Bandwidth variable optical grooming network. Routing and Wavelength and Time Assignment Integer Linear Programming (RWTA ILP) improves the Bandwidth as well as Throughput when compare to the heuristic algorithm and also in that instead of fixed length slot, variable length slot scheme (time assignment) were employed to improve the connection blocking[13]. Bandwidth Requirement in the BVOG network will be based on Bandwidth Requirement which is given by

Bandwidth Requirement = Burstlength x Capacity / time framelength

Bandwidth utilization can be improved by increasing the burst speed time and decreasing the reservation time or reservation cost [14], which is given by

Bandwidth Utilization = [T.sub.burst] / [T.sub.reservation]

Where,

[T.sub.burst] = Burst time

[T.sub.reservation] = Reservation time.

5. Throughput:

In the optical network, the most important parameter which is required for the enhanced performance of the network is the maximization of Throughput. Throughput can be maximized by many ways and also maximizing & minimizing various parameters will have impact on the Throughput.

The Throughput is highly sensitive to loss level; hence reduction in the loss (or) delay will increase (or) improve the Throughput [16]. The Throughput can be maximized by increasing the longer time slot in the transmission of packet which in turn will reduce the time slot number and latency. To succeed the above, the burst length and offset should be scheduled properly to increase the Throughput [17]. The Throughput increases with respect to random arriving traffic tolerance that can be obtained by reducing the congestion, efficient Bandwidth utilization and allocation [18].

In the Switching methodology, The optical packet switching with stop-and-wait (OPS-SAW) for the 5G network provide better reliable packet delivery than the electronic packet switching with stop-and-wait (EPS-SAW). It is likely the fact that the Optical Packet Switching stop-and-wait Throughput increases though the network load grows whereas, the Electronic packet Switching shows a degradation in the throughput when the network grows [26].

In the Ring architecture, it is observed that the maximizing Ring Rate that is the rate at which the data is sent on each span will maximize the Throughput [19]. It is given that Throughput will be high if the ring rate is double (2R) in ULSR (unidirectional line switched ring) and also 8R (n-1)/n+2 for non split 4 fibre BLSR (Bidirectional Line switched ring), where n=node of then network.

Stability is also considered as the Throughput performance metric [20], such that in the stable queue system

Ltn [right arrow] [infinity] [X.sub.ij](n)/n = Ltn [right arrow] [infinity] [A.sub.ij](n)-[U.sub.ij](n)/n = 0

Where,

[U.sub.ij]--Cumulative service applied to virtual output queue

[A.sub.ij(h)] = arrival number

[X.sub.ij(n)] = number of packet enqueued in virtual output queue

n = time.

Throughput is correlated with packet loss probability [16], such that Throughput is given by

[lambda] = 1/p+1/1-p / 1/p+[x.summation over (i=0)][[alpha].sup.i]

Where,

p = Packet loss probability

1/p = average length of successful trains

1/1-p = average length of lost trains and [alpha] = ([o.sub.11]).

It is also concluded that the increase in burst size will increase the Throughput.

6. Comparative Analysis Of All Factors:

As mentioned earlier, the prime parameters are dependant and interlinked with other parameters. A change in one parameter will cause the change in the other parameters. Hence, it is mandatory to compare their characteristics to understand their dependency. Hop Count as the parameter will be wholly depending upon the limitation of the hops from source to destination which means the limitation in the distance [2]. Hence it is the fact that, the reduction in the number of hops will reluctantly improve the propagation time and minimize the wavelength used. Since the least number of hops will encourage the path to maintain least wavelength, will indeed reduce the traffic in the particular path [4] and so, reduction in the traffic will pay a way to use the Bandwidth effectively and consequently the Throughput will be maximized.

Wavelength Assignment as a parameter will consider various constraints in assigning the wavelength for the particular path. Many Wavelength Assignment algorithms have been proposed which use the shortest path as the factor to assign the wavelength since selecting the wavelength in the shortest path will reduce the traffic and delay in transmission of data [7]. Wavelength Assignment in few uses weight function which depends on the number of link and free wavelength in the link. Few wavelength algorithms have been proposed to assign the wavelength based on the least congestion. Least congestion in the path will increase the usage of Bandwidth effectively in the path and thereby increases the Throughput.

Bandwidth the next parameter of interest encourages the allocation based on the dynamic scheme to avoid the wastage of Bandwidth [13]. Bandwidth is allocated properly based on the improvement in the burst length. Hence, increased burst length will minimize the guard time. Thereby latency can be reduced by increasing the burst length and decreasing the guard time. At the same time, the value should be maintained properly to maintain the void between the time slots and the blocking probability. Bandwidth is also a parameter which directly linked with increasing in the data rate. Hence, data rate should be maintained high to use the Bandwidth efficiently [11]. Hence, efficient Bandwidth utilization will increase the Throughput and it will reduce the latency as well as blocking probability.

Throughput improvisation concentrates on reducing the delay and loss. As already addressed that Throughput can be maximised by increasing the longer time slot with reduced guard rate. It is also imperative that the maximization of the Throughput is possible by maintaining proper burst length and offset [17]. Reducing the congestion will effectively increase the Bandwidth and also dropping the congestion rate depends on all the prime factors discussed earlier in the article. Data rate has to be improved to increase the Bandwidth, but stability should be maintained properly to balance the Throughput and loss probability. It is seen that each parameters will depend on all the other parameter of interest, it is indubitably parameters improvisation is possible only by the aggregate effect of all the other parameters. That can be obtained by means of multi objective function in the designing of the protocol.

Conclusion:

In building the future optical network, emphasis will be on architectural as well as layered level to satisfy the high Bandwidth demand. At the same time, when building the network with appropriate component, least non-linearity and dispersion should be needed. Hence, building the terabit optical network will be a challenging task. In order to provide better Quality of Service to the end user in optical network, this article addresses various performance factors by analyzing appropriate QoS parameters such as Hop Count, Wavelength Assignment, Bandwidth and Throughput. From the earlier analysis, we can understand that Hop Count, Wavelength Assignment, Bandwidth and Throughput are the prime factor in designing the network. As all the parameters of interest is concentrated in enhancing the lively performance through some extent, It is seen that Hop Count, Wavelength Assignment, Bandwidth and Throughput is concerted together successfully to receive the packet with least delay and zero tolerance to congestion. Hence, we can conclude that improvisation in the multi objective parameters will concatenate on improvisation in performance of a network through successful transmission and reception of data.

There is a wide opening of innovation in the optical network with respect to both architectural and layered level, indepth tuning and analyses can be made in the field of optical domain. So far the QoS parameter interdependency and improvisation in these parameters has been explored, it is to be emphasized and has to be analysed that how these QoS parameters can be used to balance and get along the dynamic drastically changing real time network environment.

REFERENCES

[1.] Ramaswami, R., 2006. Optical Networking Technologies: What Worked and What Didn't, IEEE Communications Mag., 44(9): 132-39.

[2.] David Peleg, Uri Pincas, 2005. Virtual path layouts optimizing total Hop Count on ATM tree networks,.

Journal of Discrete Algorithms, 3(1): 101-112.

[3.] Joan Triay, Cristina Cervell'o-Pastor, 2010. An Ant-Based Algorithm for Distributed Routing and Wavelength Assignment in Dynamic Optical Networks,.IEEE Journal on Selected Areas in Communications, 28(4): 542-552.

[4.] Venkataravikumar, P., Bachala Sathyanarayana, 2015. Optimized Segmentation based heuristic for Virtual Topology Design and its performance analysis, IJARCSMS, 3(4): 497-509.

[5.] Yousef, S. Kavian, Arash Rashedi, Ali Mahani, Zabih Ghassemlooy, 2013. Routing and Wavelength Assignment in optical networks using Artificial Bee Colony algorithm, Optik - Int. J. Light Electron Opt., 124(12): 1243-1249.

[6.] Dasari Tanuja, S. Shanawaz Basha, 2012. Hop-By-Hop Forwarding Routing Survey of Network Visualization, International Journal of Research Sciences and Advanced Engineering, 2(5): 5-11.

[7.] Shen, G., S.K. Bose, T.H. Cheng, C. Lu, T.Y. Chai, 2001. Efficient heuristic algorithms for light-path routing and Wavelength Assignment in WDM networks under dynamically varying loads, Computer Communications, 24(3): 64-373.

[8.] Afonso Jorge F. Cardoso, Joao Crisostomo W.A. Costa and Carlos Renato L. Frances, 2010. A New Proposal of an Efficient Algorithm for Routing and Wavelength Assignment in Optical Networks, Journal of Comm. and Information Systems, 25(1): 11-19.

[9.] Rajalakshmi, P. and Ashok Jhunjhunwala, 2005. Routing wavelength and time-slot reassignment algorithms for TDM based optical WDM networks, Journal Computer Communications, 30(18): 128-129.

[10.] Sathishkumar, P., V. Mahalingam, 2015. Best-Fit Wavelength Assignment Algorithm for Persistent Communication in Optical Networks, International Journal of Computer Science and Information Technologies, 6(1): 728-733.

[11.] Vipan Kumar, Kamal Deep Jangra, 2015. A probability based Dynamic Bandwidth Allocation Algorithm in Optical Network, International Journal of Advance Research in Computer Science and Management Studies, 3 (2): 71-80.

[12.] Kumarnath, J., J. Vetrimanikumar, 2013. A Survey of Routing Protocols in Optical Network," .International Journal of Engineering & Science Research, 3(1): 2266-2272.

[13.] Oliver Yul, Huan Xu, Ming Liao, 2015. Bandwidth-variable optical transport network with adjustable time-modulated and inverse-multiplexed wavelength channels, Springer, Telecomm. Syst., 60(3): 381-391.

[14.] Amit Kumar Garg and R.S. Kaler, 2008. Enhancing Bandwidth Utilization and QoS in Optical Burst Switched High-Speed Network, Journal of Microwaves, Optoelectronics and Electromagnetic Applications, 7(2): 91-100.

[15.] Chih-Ta Chiu, Yung-Chung Wang, 2013. Request-based Dynamic Bandwidth Allocation of Gigabit Passive Optical Network, Optics and Photonics Journal, 3: 165-170.

[16.] Pablo Jesus Argibay-Losada, Kseniia Nozhnina, Gokhan Sahin, Chunming Qiao Lander, 2014. Using Stop-and-Wait to Improve TCP Throughput in Fast Optical Switching (FOS) Networks over Short Physical Distances, IEEE, INFOCOM, pp: 1312-1320.

[17.] Jikai Li, Chunming Qiao, Jinhui Xu, Dahai Xu, 2007. Maximizing Throughput for Optical Burst Switching Networks. IEEE/ACM Transactions on Networking, 15(5): 1163-1176.

[18.] Pankaj Kaushik, Maninder Pal, 2012. Throughput Improvement in Coarse wavelength division Multiplexing Optical Networks, International Journal of Advanced Research in Computer and Communication Engineering, 1(4): 268-271.

[19.] Veeraraghavan, M., H. Lee J. Anderson, K.Y. Eng, 2003. A network Throughput comparison of optical metro ring architectures, IEEE Optical Fiber Communication Conference and Exhibit, A network Throughput comparison of optical metro ring architectures, pp: 763-765.

[20.] Andrew Brzezinski and Eytan Modiano, 2008. Achieving 100% Throughput in Reconfigurable Optical Networks JEEE/ACM Transactions on Networking, 16(4): 970-983.

[21.] Rajiv Ramaswami, Kumar N.Sivarajan and Galen H.Sasaki, 2010. Optical Networks, Elsevier Inc.

[22.] Thomas, E. Stern, Georgios Ellinas, Krishna Bala, 2009. Multiwavelength Optical Networks: Architectures", Design, and Control, Cambridge Control,". Cambridge university press.

[23.] Sana Tariq, 2015. Improving fairness, throughput and blocking performance for long haul and short reach optical networks, Doctoral dissertation, University of central California.

[24.] Paulo, H., G. Bezerra, Afonso J.F. Cardoso and Carlos R.L. Frances, 2010. Performance Evaluation of Algorithms for Wavelength Assignment in Optical WDM Networks, International Journal of Computer Science and Network Security, 10(1): 130-136.

[25.] Bingli Guo, Juhao Li, Yan Wang, Shanguo Huang, Zhangyuan Chen, and Yongqi He, 2013. Collision-Aware Routing and Spectrum Assignment in GMPLS-Enabled Flexible-Bandwidth Optical Network, J. Opt. Commun. Netw., 5(6): 658-666.

[26.] Pablo Jesus Argibay-Losada, Yuki Yoshida, Akihiro Maruta, and Ken-ichi Kitayama, 2016. Optical Versus Electronic Packet Switching in Delay-Sensitive 5G Networks: Myths versus Advantages, J. Opt. Commun. Netw., 8(11): 658-666.

[27.] O'Mahony, M.J., D. Simeonidou, D.K. Hunter and A. Tzanakaki, 2001. The application of optical packet switching in future communication networks, IEEE Commun. Mag., 39(3): 128-135.

(1) Kumarnath J and (2) Dr. Batri K

(1) Assistant professor, Electronic and Communication Engg. Department, PSNA College of Engineering and Technology, Dindigul, Tamilnadu, India- 624622.

(2) Associate professor, Electronic and Communication Engg. Department, PSNA College of Engineering and Technology, Dindigul, Tamilnadu, India- 624622

Received 12 May 2017; Accepted 5 July 2017; Available online 28 July 2017

Address For Correspondence:

Kumarnath J, Assistant professor, Electronic and Communication Engg. Department, PSNA College of Engineering and Technology, Dindigul, Tamilnadu, India- 624622,

E-mail: jkumarnath@gmail.com

Table I: The Variation Of Average Hyper Edge Load For Different Values Of Hop Count Average Logical Hop Count Average Load processed per node 1.18 0.18 1.5 0.5 1.87 0.87 2.11 1.11 2.74 1.74 3.04 2.04 3.68 2.68 3.86 2.86 Table II: Dropping Probability Of Just In Time (Jit), Balanced Just In Time (Bjit) And Prioritized Random Early Discard (Pred) On Us Long Haul With Different Loads Variation (W=64) Hop Dropping Probability Arriving Rate JIT BJIT PRED 1 6 0.025 0.026 0.052 3 6 0.035 0.045 0.056 5 6 0.070 0.072 0.068 8 6 0.080 0.072 0.075 Table III: Variation Of Throughput (Ff, Lu And Mu) For Different Link Count No. of links in the network Throughput(MBPS) FF LU MU 0 2 2 2 2 1.6 1.3 1.4 3 1.5 1.3 1.3 4 1.4 1.25 1.3 5 1.4 1 1.2 6 1.3 0.9 1.2 Table IV: Arrival Rate Vs Bandwidth Blocking Probability For Fsp-Ca, Hc-Mcap-Ff, Hc-Mcap-Ca And Fsp-Ff Arrival Rate Bandwidth Blocking Probability FSP-CA HC-MCAP-FF HC-MCAP-CA FSP-FF 0.1 1.00E-05 1E-4.8 1.00E-05 1E-4.8 1 1E-4.5 1E-3.9 1E-4.5 1E-3.9 10 1E-3.7 1E-2.9 1E-3.8 1E-2.8 50 1E-2.9 1E-1.9 1E-2.9 1E-1.8 100 1E-1.9 1E-0.9 1.00E-02 1E-0.8 Table V: Throughput Vs Maximum Transmission Unit (Bytes) For Ops And Eps Maximum Transmission Throughput(MB/S) Unit (Bytes) Optical packet Electronic packet switching switching 2000 225 150 4000 240 150 5200 250 150 6000 250 140 8000 250 100 9000 250 90

Printer friendly Cite/link Email Feedback | |

Author: | Kumarnath, J.; Batri, K. |
---|---|

Publication: | Advances in Natural and Applied Sciences |

Article Type: | Report |

Geographic Code: | 7IRAN |

Date: | Jul 1, 2017 |

Words: | 4365 |

Previous Article: | Comparison between ultrasonic and laser shock peening treatment on fatigue life performance of Al-Alloy 7075. |

Next Article: | Efficient cluster-based routing protocol for wireless sensor networks with mobile sink using CRSSI. |

Topics: |