The video delivery over the
Web is one of the major media delivery methods nowadays. To cooperate with the
user needs, various standardization organizations and companies are
investigating in better ways of video delivery over IP networks. The examples
of this trend of enhancement are development in codec compression, increase in
network bandwidth, proposals of efficient switching mechanisms of different
video quality schemes and more.
In this project the students
will be asked to design and implement the client side of the new advanced
video delivery method denoted HTTP Live Streaming. In this method the media is
stored on the server side as a series of files instead of one file in the
usual case. If in the usual case the server side stores the same content
encoded in different bitrates for flows with different bandwidth limitation,
in the case of HTTP Live Streaming, the server stores a different set of small
files for each of the bandwidth bounds. The latter division into small files
allows the client side (during the playback) to select a different file set
visually seamlessly according to the currently experienced network bandwidth.
+ Demonstrate the developed application playing
non-encrypted Live Streaming audio/video content
GEO-IP Database Correlation tool
Geo location capabilities are key point elements in today building
block of creating user based applications and services. Knowing user geo
location in an accurate real time mode is essential for building user
dedicated personal offering such as Ads and services.
IP address can provide such identification however its accuracy is
not guarantied as GPS.
There are several engines that provides geo location based on IP
addresses , some of them are paid and some them are free. These engines are
based in general on ISPs generosity. The cost of these engines varies from
free lite packages to paid hundreds USD per month per database.
We assume that correlating between multiple free database may give
better results. Solving collision between database by using on demand
services and operating networking protocols such as SNMP or trace routing
might give additional information which will solve the specific collision..
Building an automatic tool that crawls and correlate free
databases from the Internet.
The tool create single database where its input are which value are
agreed on all sources and where is the collision.
The collision will be examined by using additional on line queries
such as trace routing the destination , doing SNMP queries of multiple MIB o
the path to the end point and additional networking ideas provided by the
web site sources ( will be happy that more are used ):
spatial placement optimization algorithm
to the Home, a form of broadband internet connectivity based on GPON (gigabit
passive optical network) technology.
The FTTH technology
connects a OLT (optical line terminator) in a CO(central office) to ONU
(optical network unit) in households, using an optical fiber cable.
Cables are passed
on the ground through a grid of ducts which are pipes characterized by a
certain capacity (number of cables), length and connectivity with other ducts.
One fiber starting
in an OLT is split several times, using passive splitters, 1:4, 1:16, etc. up
to 32 or 64 households, reducing bandwidth, but providing service to more
The problem to
resolve is the optimal placement of splitters on a given geographical area,
minimize the use of ducts, keeping cable length constraints, according to a
Represent the duct grid (a directional graph) with constraints
such as fiber capacity, length, and spatial coordinates.
Represent the CO and households in spatial coordinates
Optimize the placement of splitters according to constraints
of total fiber length, splitting scheme and groups of houses (number of
houses), minimizing the number of cables in the OLT.
Optional: the results should be visualized using a standard
geographical map, e.g. Google Map or Google Earth.
Java, Optimization algorithms, graph theory, basic
FTTH Network: splitting schema