International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 5- May 2013
Packet Loss Control using Early Random Control at
the Network Edge
G. Srinivasa Rao*1 , Assoc. Professor & HOD, Department of Computer Science & Engineering,
Anurag Engineering College, Ananthagiri (v), Kodad (M), Nalgonda (Dt), AP, India,.
P. Gurulingam*2, student, Department of Computer Science & Engineering,
Anurag Engineering College, Ananthagiri (v), Kodad (M), Nalgonda (Dt), AP, India,
Battu Hanumantha Rao*3, Assoc. Professor & Dean-Academics,
N.H. College of Engineering, Parli Vaijanath, Beed District, Maharashtra – 431515, India
Abstract— This paper discusses the use of link-sharing
mechanisms in packet networks and presents algorithms for
hierarchical link-sharing. Hierarchical link-sharing allows
multiple agencies, protocol families, or traffic types to share the
bandwidth on a link in a controlled fashion. This paper presents
Random Early Detection (RED) gateways for congestion
avoidance in packets witched networks. The gateway detects
incipient congestion by computing the average queue size. The
gateway could notify connections of congestion either by
dropping packets arriving at the gateway or by setting a bit in
packet headers. We propose fair adaptive bandwidth allocation
(FABA), a buffer management discipline that ensures a fair
bandwidth allocation amongst competing flows even in the
presence of non-adaptive traffic. The RED gateway has no bias
against bursty traffic and avoids the global synchronization of
many connections decreasing their window at the same time.
Simulations of a TCP/IP network are used to illustrate the
performance of RED gateways.
coupled with gateway scheduling algorithms that require perconnection state in the gateway. end-to-end delay, as well as
from packet drops or other methods. Nevertheless, the view of
an individual connection is limited by the timescales of the
connection, the traffic pattern of the connection, the lack of
knowledge of the number of congested gateways, the
possibilities of routing changes, as well as by other difficulties
in distinguishing propagation delay from persistent queueing
delay. The most effective detection of congestion can occur in
the gateway itself.
Keywords— Bandwidth, Time Complexity, Space Complexity,
Active flows.
I. INTRODUCTION
In high-speed networks with connections with large delaybandwidth products, gateways are likely to be designed with
correspondingly large maximum queues to accommodate
transient congestion. In the current Internet, the TCP transport
protocol detects congestion only after a packet has been
dropped at the gateway. However, it would clearly be
undesirable to have large queues (possibly on the order of a
delay-bandwidth product) that were full much of the time; this
would significantly increase the average delay in the network.
Therefore, with increasingly high-speed networks, it is
increasingly important to have mechanisms that keep
throughput high but average queue sizes low. In the absence
of explicit feedback from the gateway, there are a number of
mechanisms that have been proposed for transport-layer
protocols to maintain high throughput and low delay in the
network. Some of these proposed mechanisms are designed to
work with current gateways while other mechanisms are
ISSN: 2231-5381
Figure 1: System Architecture.
The gateway can reliably distinguish between propagation
delay and persistent queueing delay. Only the gateway has a
unified view of the queueing behavior over time; the
perspective of individual connections is limited by the packet
arrival patterns for those connections. In addition, a gateway
is shared by many active connections with a wide range of
roundtrip times, tolerances of delay, throughput requirements,
etc.; decisions about the duration and magnitude of transient
congestion to be allowed at the gateway are best made by the
gateway itself. The method of monitoring the average queue
size at the gateway, and of notifying connections of incipient
congestion, is based of the assumption that it will continue to
http://www.ijettjournal.org
Page 1954
International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 5- May 2013
be useful to have queues at the gateway where traffic from a
number of connections is multiplexed together, with FIFO
scheduling. Not only is FIFO scheduling useful for sharing
delay among connections, reducing delay for a particular
connection during its periods of burstiness [4], but it scales
well and is easy to implement efficiently. In an alternate
approach, some congestion control mechanisms that use
variants of Fair Queueing [20] or hop-by-hop flow control
schemes [22] propose that the gateway scheduling algorithm
make use of per-connection state for every active connection.
We would suggest instead that per-connection gateway
mechanisms should be used only in those circumstances
where gateway scheduling mechanisms without perconnection mechanisms are clearly inadequate.
Requirements for resource management in the Internet
include both services for real-time traffic and link-sharing
services. Realtime traffic is characterized by a (fixed or
adaptive) playback time at the receiver; real-time packets
arriving at the receiver after the playback time are discarded.
In a congested network, resource management mechanisms
are required at the gateway to meet realtime traffic
requirements for controlled delay and limited packet drops.
While there has been an abundance of research about the
needs of real-time traffic, link-sharing services have received
somewhat less attention in the research community. The
approach to controlled link-sharing described in this paper has
evolved in the context of the Internet. Because the Internet is
decentralized in nature, composed of multiple administrative
domains with a wide range of resource limitations, the control
of Internet resources involves local decisions on usage as well
as considerations of per-connection end-to-end requirements.
One function of link-sharing mechanisms is to enable
gateways to control the distribution of bandwidth on local
links in response to purely local needs. By allowing isolation
between real-time and best-effort traffic in cooperation with
packet scheduling algorithms that give priority to the real-time
traffic, controlled link-sharing can also be a key component in
enabling the deployment of priority-based packet scheduling
algorithms designed to meet the end-to-end service
requirements of realtime traffic.
II. BACKGROUND WORK
The basic idea of P2P network is to have peers participate
in an application level overlay network and operate as both A
number of approaches for queue management at Internet
gateways have been studied earlier. Droptail gateways are
used almost universally in the current Internet due to their
simplicity. A droptail gateway drops an incoming packet only
when the buffer is full, thus providing congestion notification
to protocols like TCP. While simple to implement, it
distributes losses among flows arbitrarily[9]. This often results
in bursty losses from a single TCP connection, thereby
reducing its window sharply. Thus, the flow rate and
consequently the throughput for that flow drops. Tail dropping
also results in multiple connections simultaneously suffering
losses leading to global synchronization [10]. RED addresses
some of the drawbacks of droptail gateways. An RED
ISSN: 2231-5381
gateway drops incoming packets with a dynamically
computed probability when the exponential weighted moving
average queue size avg q exceeds a threshold called min th.
This probability increases linearly with increasing avg q till
max th after which all packets are dropped until the avg q
again drops below max th. The RED drop probability also
depends on the number of packets enqueued since the last
packet drop. The goal of RED is to drop packets from each
flow in proportion to the amount of bandwidth it is using.
However, from each connection _s point of view the packet
loss rate during a short period is independent of the bandwidth
usage as shown in [10]. This contributes to unfair link sharing
in the following ways:
• Even a low bandwidth TCP connection observes packet
loss which prevents it from using
its fair share of bandwidth.
• A non-adaptive flow can increase the drop probability of
all the other flows by sending at a fast rate, which increases
the drop probability for all flows.
• The calculation of avg q for every packet arrival is
computationally intensive.
To ensure fairness amongst flows in terms of bandwidth
received and to identify and penalize misbehaving users, it is
necessary to maintain some sort of per-flow state [14]. Many
approaches based on per-flow accounting have been suggested
earlier. FRED [10] does per-flow accounting maintaining only
a single queue. It suggests changes to the RED algorithm to
ensure fairness and to penalize misbehaving flows. It puts a
limit on the number of packets a flow can have in the queue.
Besides it maintains the per flow queue occupancy. Drop or
accept decision for an incoming packet is then based on
average queue length and the state of that flow. It also keeps
track of the flows which consistently violate the limit
requirement by maintaining a per-flow variable called strike
and penalizes those flows which have a high value for strike.
It is intended that this variable will become high for nonadaptive flows and so they will be penalized aggressively. It
has been shown through simulations [12] that FRED fails to
ensure fairness in many cases. CHOKe [13] is an extension to
RED. It does not maintain any per flow state and works on the
good heuristic that a flow sending at a high rate is likely to
have more packets in the queue during the time of congestion.
It decides to drop a packet during congestion if in a random
toss; it finds another packet of the same flow.
In a significant paper by Guerin et al. [19], the authors
establish how rate guarantees can be provided by simply using
buffer management. They show that the buffer management
approach is indeed capable of providing reasonably accurate
rate guarantees and fair distribution of excess resources.
III. METHODS
FABA algorithm
If the buffer size of the bottleneck router is B, then it is
easy to see that the space complexity of FABA algorithm is.
This can be argued as follows: the number of token buckets is
equal to the number of active flows passing through the router
and in the worst case, there are B active flows at the
http://www.ijettjournal.org
Page 1955
International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 5- May 2013
bottleneck buffer. We have already argued that by the nature
of token distribution, if a packet is in position x from the front
of the queue, the average number of tokens added to its bucket
when the packet is dequeued is x=N.
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
Figure 2: Simulation Setup
[11]
[12]
[13]
[14]
[15]
[16]
[17]
Figure 3: Packet Loss control using FABA
[18]
[19]
IV. CONCLUSION
[20]
The primary advantage of the probing mechanism can be
understood as follows. For normal messages without the
probing mechanism, the source does not have any idea about
the status of the destination. So fi the request to connect is lost,
the entire process is in jeopardy. In messages using the
probing mechanism, the request is sent only if the status of the
destination is optimal for transactions to being. Therefore, the
entire process is more stabilized in the previous version. One
of the defects in this model is the use of static routing and the
model can be improved by implementing dynamic routing.
Using Java we created a sample intranet structure to simulate
probing mechanism. After the network topology has been
chosen, the parameters necessary to execute it are supplied to
the simulator. Likewise various network structures can be
created using this tool and various congestion detecting and
control mechanisms can be simulated.
[21]
[22]
[23]
[24]
[25]
[26]
[27]
REFERENCES
[1]
[2]
L.L. Peterson, B.S. Davie, Computer Networks: A Systems Approach,
Morgan Kaufmann, Los Altos, CA, 1996.
D.C. Verma, Policy-Based Networking: Architecture and Algorithms,
New Riders Publishing, Indianapolis, IN, USA, 2000.
ISSN: 2231-5381
[28]
S. Keshav, An Engineering Approach to Computer Networking: ATM
Networks, the Internet, and the Telephone Network, Addison-Wesley,
Reading, MA, 1997.
D. Bertsekas and R. Gallager. Data Networks. Prentice Hall,
Englewood Cliffs, NJ, 1987.
P. Billingsley. Convergence of Probability Measures. Wiley, 1968.
F. Clevenot and P. Nain. A Simple Fluid Model for the Analysis of the
Squirrel Peer-to-Peer Caching System. In Proceedings of IEEE
INFOCOM, 2004.
F. Clevenot, P. Nain, and K. Ross. Stochastic Fluid Models for Cache
Clusters. Technical Report 4815, INRIA, Sophia Antipolis, 2003. To
appear in Performance Evaluation.
B. Cohen. Incentives build robustness in bittorrent, May 2003.
http://bitconjurer.org/BitTorrent /bittorrentecon.pdf.
F. Dabek, M. F. Kaashoek, D. Karger, R. Morris, and I. Stoica. Widearea cooperative storage with CFS. In Proceedings of the 18th ACM
Symposium on Operating Systems Principles (SOSP ’01), Chateau
Lake Louise, Banff, Canada, October 2001.
A. Das and R. Srikant. Diffusion approximations for a single node
accessed by congestion-controlled sources. IEEE Transactions on
Automatic Control, 45(10):1783–1799, October 1998.
G. de Veciana and X. Yang. Fairness, incentives and performance in
peer-to-peer networks. In the Forty-first Annual Allerton Conference
on Communication, Control and Computing, Monticello, IL, Oct. 2003.
Sally Floyd, Van Jacobson, Link-sharing and Resource Management
Models for Packet Networks, IEEE\ACM Transactions on Networking,
Vol.3, No.4, 1995.
John Nagle, RFC896 congestion collapse, January 1984.
Sally Floyd and Kevin Fall, Promoting the Use of End-to-End Congestion
Control in the Internet, IEEE/ACM Transactions on Networking, August
1999.
V. Jacobson. “Congestion Avoidance and Control”. SIGCOMM
Symposium on Communications Architectures and Protocols, pages 314–
329, 1988.
http://www.isi.edu/nsnam/ns/
L. Benmohamed and S. M. Meerkov, “Feedback control of congestion in
packet switching networks: The case of a single congested node,”
IEEE/ACM Trans. Networking, vol. 1, pp. 693–708, Dec. 1993.
E. Altman, T. Basar, and R. Srikant, “Control methods for communication networks,” in Proc. 36th Conf. Decision and Control, San Diego, CA,
1997, pp. TA 31774–1809, TM3 2368–2404, TP3 2903–2945.
S. H. Low and D. E. Lapsey, “Optimization flow control—I: Basic algorithms and convergence,” IEEE/ACM Trans. Networking, vol. 7, pp.
861–874, Dec. 1999.
S. Floyd and K. Fall, “Promoting the use of end-to-end congestion con- trol
in the Internet,” IEEE/ACM Trans. Networking, vol. 7, pp.458–472, Aug.
1999.
S. J. Golestani and S. Bhattacharyya, “A class of end-to-end congestioncontrol algorithms for the Internet,” in Proc. Int. Conf. Network Protocols
(ICNP), Austin, TX, Oct. 1998, pp. 137–150.
A. Kolarov and G. Ramamurthy, “A control theoretic Approach to the
design of explicit rate controller for ABR service,” IEEE/ACM Trans.
Networking, vol. 7, pp. 781–753, Oct. 1999.
N. Ghani and J. W. Mark, “Enhanced distributed explicit rate allocation for
ABR services in ATM networks,” IEEE/ACM Trans. Networking, vol. 7,
pp. 710–723, Oct. 1999.
S. Kalyanaraman, R. Jain, R. Goyal, S. Fahmy, and B. Vandalore, “The
ERICA switch algorithm for ABR traffic management in ATM networks,”
IEEE/ACM Trans. Networking, vol. 8, pp. 87–98, Feb. 2000.
T. V. Lakshman, P. P. Mishra, and K. K. Ramakrishan, “Transporting
compressed video over ATM networks with explicit-rate feedback control,”
IEEE/ACM Trans. Networking, vol. 8, pp. 71–86, Feb. 2000.
R. Muthukrishnan, S. Dasgupta, A. Varma, L. Kalampoukas, and K. K.
Ramakrishnan, “Design, implementation and evaluation of an explicit rate
allocation algorithm in an ATM switch,” in Proc. IEEE INFOCOM, Tel
Aviv, Israel, Apr. 2000, pp. 1313–1322.
Y. Zhao, S. Q. Li, and S. Sigarto, “A linear dynamic model for design of
stable explicit-rate ABR control schemes,” in Proc. IEEE INFOCOM,
Kobe, Japan, Apr. 1997, pp. 283–292.
S. Mascolo, D. Cavendish, and M. Gerla, “ATM rate-based congestion
control using a Smith predictor: An EPRCA implementation,” in Proc.
IEEE INFOCOM’96, San Francisco, CA, 1996, pp. 569–576.
http://www.ijettjournal.org
Page 1956
International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 5- May 2013
[1] G.Srinivasa Rao is currently
working as an Associate Professor
& Head of the Department of
Computer Science & Engineering
at Anurag Engineering College,
Kodad. He guided many projects
in the area of image processing for
CSE & IT Departments. His
research intrests are at the areas of
Data Warehousing and Network
Security.
[2] P.Gurulingam is pursuing
M.Tech(CSE) at the Department
of
Computer
Science
&
Engineering , Anurag Engineering
College Kodad. He received
B.Tech CSE from SV Univeristy,
Tirupati. His research intrests are
at the area of Computer Networks.
[3] B.Hanumantha Rao is
currently
working
as
an
Associate Professor at the
Department of Computer Science
& Engineering, and also DeanAcademics, NH College of
Engineering. He is also a
research
scholar
at
the
Department of Computer Science and Engineering,
Acharya Nagarjuna University,Guntu. He completed his
M.Sc (Computer Scinece), M.Tech(CSE) and MBA ..His
research work focouses on Software Development
Methodologies, and also he intrests at the areas of
Computer Networks and Operating Systems.
ISSN: 2231-5381
http://www.ijettjournal.org
Page 1957