Multimedia Design Issues for Internet Telephony Protocols in Current High Performance Networks by ijcsiseditor


									                                      (IJCSIS) International Journal of Computer Science and Information Security,
                                      Vol. 9, No. 5, May 2011


A.Jayachandran                  Dr.R.Dhanasekaran,Principal               P. Rajan, Professor in MCA
Asst Prof,CSE Dept                                                       dept, PSN College of
PSN College of Engineering       Syed Ammal Engineering
                                college,Ramnad,INDIA                     Engineering and Technology,
and technology

                                                        1. INTRODUCTION
The rapid growth of media content
distribution[5] on the internet in the past          We describe how a scalable system of dense
few years has brought with it                        Wi-Fi     sensors[6,7]    can    be    built
commensurate increases in the cost of                inexpensively. We build such a system and
distributing the content. Underlying the             evaluate its performance. We provide
internet access trends at a global scheme            specific example of why standard
i.e. how people use the internet is a                authentication and encryption scheme are in
challenging     problem     i.e.  typically          adequate to secure corporate Wi-Fi networks
addressed by network frames. We seek to              which motivate our solutions based on
understand the intrinsic reasons for the             continuous monitoring of Wi-Fi networks.
well known phenomenon of heavy tailed                We show that, to provide comprehensive
degree in the internet. AS graph and                 coverage for detecting security branches a
anywhere that the contrast to traditional            dense deployment of RF sensor is necessary.
model based on preferential attachment               Magnetic and electromagnetic sensors do
and centralized optimization the present             not require direct physical content and are
degree of the internet can be explained by           useful for detecting proximity effects.
the evolution of wealth associated with              Magneto resistive effect is a related
each ISP. Our extensive numerical and                phenomenon depending on the fact that the
simulation examples have shown that the              conductivity varies as the square of the
proposed scheme achieves satisfied                   applied flux density. Magnetic field sensors
accuracy and computational efficiency.               can be used to detect the remote presence of
The effectiveness of the proposed                    metallic objects. Eddy current sensors use
detection and trace back methods are                 magnetic probe coils to detect in the metallic
verified through extensive simulations               structure such as pipes.
and internet datasets.
                                                     Thermal sensors are a family of sensors used
Key words : Peer to peer , Network                   to measure temperature or heat flux. Most
security,  Security policy, Content                  biological organisms have developed
delivery, 3G/4G networks, Grid and                   sophisticated temperature sensing systems.
Cloud computing                                      Thermo resistive effects are based on the

                                                                                 ISSN 1947-5500
                                          (IJCSIS) International Journal of Computer Science and Information Security,
                                          Vol. 9, No. 5, May 2011

fact that the resistance R changes with                     3. Caching and content routing service
temperature T. For moderate changes the                        capacity
relation is approximately given for many
                                                                   Multicast over the internet was
metals by ∆R/R = R ∆T with R the
                                                                   originally proposed at the network
temperature coefficient of resistance. The
                                                                   layer referred to as multicast.
relationship for silicon is more complicated
                                                                   However after a decade of research
but it is well understood. Hence silicon is
                                                                   there are still many hurdles in the
useful for detecting temperature changes.
                                                                   deployment of IP multicast such as
Resonant temperature sensors rely on the                           the    lack     of   higher    layer
fact that single crystal sio2 exhibit a change                     functionalities and scalable inter
in resonant frequency depending on                                 domain multicast routing protocols.
temperature change. Since this is a
                                                                   Data management – A document
frequency effect it is more accurate than
                                                                   may be portioned into various parts
amplitude change effect and has extreme
                                                                   permitting concurrent downloading
sensitivity and accuracy for small
                                                                   from multiple peers[1] , the
temperature changes.
                                                                   granularity and placement of these is
 2. On demand multimedia content                                   critical.

For a large content delivery networks that                         Peer selection – The mechanism
consists of hundreds are even thousands of                         whereby a peer is selected as a server
geographically distributed CDN[5] servers                          may take into account load balancing
like the hierarchical topology of the IP                           bandwidth       availability      and
routers on the internet we believe that a                          differentiate among peers who
hierarchical overlay network topology is                           contribute more to the community.
required for the CDN servers to perform
                                                                   Admission and scheduling policy –
content routing and content delivery done
                                                                   Limiting the number of concurrent
                                                                   down loaders and/or scheduling to
Step 1: Try to satisfy the    users    request                     provide    differentiation priority
using local CDN server.                                            among them.

Step 2: If step 1 fails try to satisfy the user                    Traffic – The request processes for
request using a CDN server inside the                              documents along with the dynamic
cluster including the local CDN server                             of how peers stay online and/or
                                                                   delete documents.
Step 3: If step 2 fails try to satisfy the user
request using a CDN server inside a nearby

Step 4: If step 3 fails try to satisfy the user
request using the origin server

                                                                                     ISSN 1947-5500
                                      (IJCSIS) International Journal of Computer Science and Information Security,
                                      Vol. 9, No. 5, May 2011

4. Methodology                                                 (c) Table comparisons

    (a) Service capacity – since BT uses                       Existing:
        multiport downloads service
        capacity must be carefully                             Capacity          CD delay           P2P delay
        defined. We estimate the service                       1.2626            0.2619             0.2676
        capacity as
                                                               0.2159            0.2261             4.7244
    Effective number of service = Total
    storage space / size                                       0.1960            0.1960             0

    (b) The     effective    number      of                    0.1757            0.1968             12.0091
        replicates available in the system
        including partial downloads.                           0.1480            0.0.1472           0.5405
        Since peers may exist or delete a                      0.1360            1.1371             0.0968
        file upon completing a download
        we can estimate service capacity
        based on the following formula
    T * Volume – ( Number of finished –
    Number of seeds ) * size / size                            Capacity           CD delay              P2P
    (c)Throughput and delay of each
    peer                                                       0.13481            0.13472               0.0668

            We estimate the average                            0.11180            0.11675               4.4275
    instantaneous throughput seen by
                                                               0.09929            0.10119               1.9136
    each peer as follows
                                                               0.066592           0.06587               0.0758
    Average throughput per peer =
    throughput / number of downloads                           0.03264            0.03221               1.3174

                                                               0.01649            0.01575               4.4876

                                                               We describe a set of network
                                                               benchmark for measuring bandwidth
                                                               , latency , software overhead which
                                                               we have implemented over a wide
                                                               variety of networks. We provide data
                                                               from these benchmarks for both
                                                               small      and     large    message
                                                               performance on many of these

                                                                                 ISSN 1947-5500
                                      (IJCSIS) International Journal of Computer Science and Information Security,
                                      Vol. 9, No. 5, May 2011

    benchmark for both small and large                  6. Mobility management                         in       4G
    message performance on many of the                     networks using IPV6
    super computer networks in use
    today and compare the performance                          High speed – 4G[9,10] systems
    of MPI to that of lower level network                      should offer a peak speed of more
    API.                                                       than 10 Mbits per second in
                                                               stationery mode with an average of
    Using our results, we examine                              20 Mbits per second when traveling.
    various application speedups that can
    be achieved via network related                            High network capacity – It should be
    optimizations such as overloading                          at least 10 times that of 3G[9,10]
    computations with communications,                          systems. This will quicken the time
    pipelining messages and the use of                         load time of 10 MB file to 1 second
    message packing. We provide a                              on 4G from 200 seconds on 3G
    historical portrait of the trends in                       enabling high speed video to stream
    small message performance over the                         to phone and create a virtual reality
    past 10 years.                                             experience on high resolution hand
                                                               set screens fast/seamless handover
5. Resource reservations                                       across multiple networks. 4G
                                                               wireless networks[8,11] should
    The site manager responsible for a                         support global roaming across
    resource makes a donation to a                             multiple wireless and mobile
    resource broker of its choosing for an                     networks.
    agreed upon number of tokens.
    When a user wants to ‘buy’ this                            Next    generation   multimedia[4]
    resource it contact the responsible                        support – The underlying network
    resource broker and pays the                               for 4G must be able to support fast
    required number of token and when                          speed and large volume of data
    the tokens ‘validity’ is checked get a                     transmission at a lower cost than
    ticket for that resource in return. The                    today.
    ticket is presented to the component
    manager responsible for the resource                       128 bit address space provides a
    which verifies it a authenticity. The                      sufficiently large number of address.
    component manager then gives the                           High quality support for real time
    user access credentials for that                           audio transmission short/ bursty
    resource. The network hierarchy has                        connections of web applications ,
    3 main hierarchical components i.e.                        peer to peer[1] applications etc.,
    user equipment, Radio access                               Faster packet delivery , decreased
    network and core network.                                  cost processing no header checksum
                                                               at each relay fragmentation only at
                                                               the end points. Smooth hand off
                                                               when the mobile host travel from one

                                                                                 ISSN 1947-5500
                                      (IJCSIS) International Journal of Computer Science and Information Security,
                                      Vol. 9, No. 5, May 2011

    subnet to another. Causing a change                                                 Buffer
    in its care of address.

7. Range of wireless services and
   address allocations
    Broadcasting services – It include
                                                                    IP                                         IP Address
    short waves like AM and FM radio
    as well as terrestrial television.                              Addres

    Mobile communication of voice and
    data – including maritime and
    aeronautical         mobile         for
    communications between ship ,                                                                                  X.25 packet
                                                               Token Ring
    airplanes and land mobile for                                                                                   switched
    communications between a fixed
    base station and moving size such as                                            Fig 1.1
    a taxi fleet and paging services and
                                                               This another case that seems to
    mobile communications. Either
                                                               duplicate lower layer features. Some
    between mobile user and a fixed
                                                               errors do escape lower layer error
    network or between mobile user such
                                                               detection. This statement may sound
    as mobile telephone services. Fixed
                                                               unusual because it implies that even
    services either point to point or point
                                                               if data link error detection provides
    to multi point services. Satellite used
                                                               reliable transmission along each link
    for broadcasting telecommunications
                                                               there is still no guarantee of error
    and internet particularly over long
                                                               free transmission between the source
    distances. Other user including
                                                               and destination. Suppose it receives
    military        radio       astronomy
                                                               the IP packet intact but an error that
    metrological or scientific users.
                                                               affect the packet current occur
                                                               during reformatting of the frame
                    Client                                     containing the packet.


                   Interface                                           Transport
                                                                                             Fig 1.2



                                                                                 ISSN 1947-5500
                                    (IJCSIS) International Journal of Computer Science and Information Security,
                                    Vol. 9, No. 5, May 2011

A client or server application interact                      address
directly with a transport layer
protocol to establish common and to                          Un assigned          011             1/8
send or receive information. The                             Reserved for 100                     1/8
transport layer protocol then uses                           geographic
lower layer protocol to send or                              based
receive individual messages , then a                         unicast
computer needs a complete stack of                           address
protocols to run either a client or a
server.                                                      Un assigned          101             1 / 18

Allocation      Prefix     Fraction of                       Un assigned          110             1/8
space                      address
                           space                             Un assigned          1110            1 / 16

Reserved        0000       1 / 256                           Un assigned          1111 0          1 / 32
                0000                                         Un assigned          1111 10         1 / 64
Un assigned     0000       1 / 256                           Un assigned          1111            1 / 128
                0001                                                              110
Reserved for 0000          1 / 128                           Un assigned          1111            1 / 512
NSAP         001                                                                  1110 0
                                                             Link local 1111                      1 / 1024
Reserved for 000 010       1 /128                            use address 1110 10
allocation                                                   Size    local 1111                   1 / 1024
                                                             use address   1110 11
Un assigned     0000       1 / 128
                011                                          Multicast            1111            1 / 256
                                                             address              1111
Un assigned     0000 1     1 / 32

Un assigned     0001       1 / 16

Un assigned     001        1/8

Provider        010        1/8

                                                                               ISSN 1947-5500
                                      (IJCSIS) International Journal of Computer Science and Information Security,
                                      Vol. 9, No. 5, May 2011

8. Strategic instruments        to   limit                     for termination. Such access prices
   competition                                                 are often negotiated privately
                                                               between mobile operators.
    A call from a mobile operator to a
    fixed line network involves the                            Wayne Thomson – SAS has been
    mobile operator collecting all the                         one of the first vendors to recognize
    revenue for the call but the fixed line                    the need for operationalizing
    network providing the termination of                       analytics i.e. moving beyond just
    the call. As a result, the fixed line                      knowledge discovery. Open standard
    network usually changes for the                            have been very critical to deploy
    termination of mobile to fixed calls.                      analytical models in a heterogeneous
    Typically the termination of mobile                        environment.
    to fixed call is treated in the same
    way as the termination of other calls                      Mike Hockins – We will be
    to fixed line networks being                               announcing our data rush based
    regulated on a cost basis.                                 parallel data mining solution for the
                                                               summer. Shortly thereafter we
    A call from a fixed line network to a                      expect to announce support for
    mobile operator often leads to an                          reading and writing PMML both at
    access price being set by the mobile                       design time and run time.
    operator. In the majority of market
    the termination of fixed to mobile                         Robert Grossmann – I think cloud
    calls has been priced many times                           computing[10] is the disruptive
    higher than the termination of mobile                      technology that will have the greatest
    to fixed calls, some times resulting in                    impact on data mining and predictive
    the regulation of fixed to mobile                          model in the near term.
    termination process. An exception is                9. Metrological and scientific issues in
    the US where this access price is the                  internet
    same as the mobile to fixed access
    price due to the symmetry                                  Processor in the cluster may not be
    requirements under the 1996 telecom                        identical.   The     communication
    act.[2,3] Other exceptions are                             network may have a regular but
    Canada, Hong Kong and Singapore                            heterogeneous structure. The cluster
    where there are no fixed to mobile                         may be a multi-user computer
    termination changes. Instead like the                      system. This is in particular makes
    US mobile operators[4] recover the                         the performance character of the
    cost of terminating fixed to mobile                        processor dynamic and non identical.
    calls from the mobile callers directly.                    Mobile telecom system with
                                                               different type of processor from
    A call from a mobile operator to                           embedded into mobile phones to
    another mobile operator may involve                        central computers processing calls.
    mobile operator charging each other

                                                                                 ISSN 1947-5500
                                 (IJCSIS) International Journal of Computer Science and Information Security,
                                 Vol. 9, No. 5, May 2011

Embedded control multiprocessor                           paging all processors are supposed to
system ex. Car, airplanes etc., Given                     demonstrate the same speed. Each
a set of processors the speed of each                     processor is characterized by the two
of which is characterized by a                            parameters i.e. Current CPU
positive constant.                                        utilizations and the size of current
                                                          available memory.
Partition is the mathematical object
into sub objects of the same type.                        The serial code is provided by the
There is one to one mapping between                       application programmer. It is
the position and the processors. The                      supposed      that   the    code is
size of each partition is proportional                    representative for the computations
to the speed of the processor owing                       performed during the execution of
the partition. Some additional                            the application. The code is
restrictions to the relationship                          performed at runtime in the points on
between the partitions are satisfied.                     the application specified by the
The partitioning minimizes some                           application programmer. It is
functional, which is used to measure                      supposed      that   the    code is
each partitioning. The whole                              representative for the computations
computation is partitioned into a                         performed during the execution of
number of equal chunks. Each chunk                        the application. The code is
is performed by separate process.                         performed at runtime in the points of
The number of process run by each                         the application specified by the
processor is proportional to the                          application programmer. Thus the
relative speed of the processor.                          performed model of the processor
                                                          provides current estimation of their
The performance of heterogeneous                          speed demonstrated on the code
model network of computers                                representative for the particular
quantifying the ability of the network                    application. The number of process
to perform computations and                               executing the algorithm which
communications. Performance model                         normally of the parameter         the
of applications quantifying the                           model. The total volume of the
computations and communications to                        computations to be performed by
be performed by the application.                          each processor during the execution
Performance model of a set of                             of the algorithm. The total volume of
heterogeneous processor which is                          data transferred between each pair of
used to estimate the execution time                       the process during the execution of
of computations. Performance model                        the algorithm.
of communication network which is
necessary to predict the execution                        Currently owned and used space
time of communication operations.                         from the data user ISP’s. New space
In the operation of external load and                     obtained by directly from internet

                                                                            ISSN 1947-5500
                                         (IJCSIS) International Journal of Computer Science and Information Security,
                                         Vol. 9, No. 5, May 2011

       registries. Customer space private               API – Application program interface
       autonomous system currently owned
       and used data center. A new AS                   SAS – Source address service
       obtained directly from customer                  PMMI – Phase modulation mobile interface
       internet registries. One to one, One
       to many, Many to one and Many to                 NAT – Network address technology
       many relationships are possible. 25%
                                                        VNT – Virtual network terminal
       of the new space must be utilized
       immediately and 50% of the space is              References
       used within one year. To get more
       space and the added space must be                     1. Analysis of peer to peer internet
       utilized as 80%. The private IP                          telephony protocols by Salman A
       address conflict with other customer                     Basket
       networks. When the network
                                                             2. Cyber security threats and counter
       communication is initiated and
                                                                measures by Buheita Fujiwara
       where the application that need to be
       run and that can be work across NAT                   3. Network security policy by Frederic
       boundaries.[10,11]                                       M Avolio July 2007

Conclusion                                                   4. Multimedia      content    delivery
                                                                networks by Jian Ni and H K Tsang
         In this paper we presented an
alternative theory of the internet evolution                 5. Evaluation             of   current              high
and developed a new topology generator                          performance            networks by               Paul
based on wealth evolution and random                            Hargrove
walks. Our method reconfigures the VNT
gradually by dividing it into multiple stages.               6. Enhancing the security of Wi-Fi
This would reduce dramatically the sites of                     networks by Paramvir Bahl
matrices whose pseudo inverse matrices                       7. Wireless sensor networks by F L
needed to be calculated at each stage. The                      lewis
eventual goal is the practical design of each
scheme to achieve the performance outlined                   8. Flexible resource allocation in 3G
in this paper.                                                  networks by Mohamed Al-Fares

Abbreviations                                                9. Wireless communications by Joshua
                                                                S Gans
ISP – Internet service Protocol
                                                             10. Open standard and cloud computing
RFC – Requested flow control                                     by Michael Zeller Zementics
CDN – Content delivery networks                              11. High performance grid computing by
MPI – Multi Program Interface                                    Jack Dongarra

                                                                                    ISSN 1947-5500

To top