Category: CSE Seminar Topics

Introduction to Quality of Service Routing Approaches and Algorithms Seminar Topic:

It is used to define the service class that a packet should receive at the end of each switch. It also used to allocate certain amount of resources. It also sorts the incoming packets to their respective classes. It controls the amount of traffic to the each class. These steps are followed again to each and every switch, or at least all bottleneck routers.

Why we use QOS routing. It is basically used for dynamic determination of feasible paths. It also optimizes the resource usage. And improve the performance degradation.

Routing strategies and approaches: it includes source routing for feasible paths that computed by the source node.

 In distributed routing they have paths that are computed by distributed computation.

 In hierarchical routingtopology are created by clustering nodes into groups that form logical nodes.

Heuristic approach: it uses famous Chen and nahrstedts algorithm, and jaffe’s algorithm that proposes non-linear path length function.

 They use three metrics: available bandwidth, propagation delay, hop count. For selecting path they use the technique of bellman-ford pre-computation, it computes the bandwidth for all counts. They also use Dijkstra on demand and Dijkstra pre-computation. For forwarding purpose they use hop-by-hop method.

It has some problems also that arises during QoS routing: it creates routing with in accurate information. Networks are dynamic and have aggregation problem also. Information is hidden and the calculation is not approximate. Ithas some inter-class effects in QoS routing. It uses bandwidth reservation to prevent the starvation of low priority traffic.

Conclusion:it is require finding some new path for sufficient resources available. QoS traffic reserve resources while major of them do not. User requirement are successful, and satisfied. They use shortest path algorithms like dijkstra algorithm and bellman-ford algorithm.

 Download Technical Seminar on Quality of Service Routing Approaches and Algorithms PPT .

Introduction to Web mining Seminar Topic:

Web mining id divided into three parts: web content mining: it is used for unstructured data semi-structured data generally for html, web structure mining: it is used for hyperlinks on the web as an information source and web usage mining: used for interaction with the web server.

Web content mining: it is used for use for useful data from web like document, image, audio, video, metadata and hyperlinks. It is to improve the information finding and filtering the user information on the web. Integrate the data base on the web such that it works for any query.

Web structure mining:  it used for the hyperlink ion the webpage that are used to generate the pages of documents and structural summary of the inter-connected document. It is based on three types: first: it is based on the hyperlinks, pages and generated information. Second: using the web document itself. Third: discovering the network of hyperlinks in the webpage.

Web usage mining: It is also called web mining log. It used to make site better. When user is using that web page then they interact with usage pattern with the interaction of user. It improves the quality of information transfer to user through internet.

Webpages work as nodes and the hyperlink on the pages work as edges.  A web page that contains the lot of information about the relevant topic then it is called as authorities’ and the page that contains large number of hyperlinks about any topic then it hubs.

Web data clustering:  it used to make web page simple and easy access like it reduces the length navigation path, improve content delivery, information retrieval system. They are of many types like hierarchical, partition, graph based, fuzzy, hybrid.

Web caching: it improves the net traffic and server load. It also reduces bandwidth consumption.

 Download Technical Seminar on Web Mining Taxonomy PPT .

Introduction to Sub Netting and VLSM Seminar Topic:

IP address:  it is a 32 bit binary number represented in 4 decimal values each represent 8 bits range from 0 to 255. It holds the address of network id and host id. It is also called internet protocol address. It is divided into 5 class a, b, c, d and e.

Class A begins with 0xxx. It has value in range 1 to 126. Class B begins 10xx. It has range from 128 to 191. Class C begins with 110x. It has range from 192 to 239. Class D begins with 1110. It has range from 224 to 239. Class E begins from 1111. It has range from 240 to 254. In which CLASS A, B and C are most widely used. Class d is used for addresses and class e is used for future reference.

For every network address host section is set to all 0’s and during broadcasting it is set to 1.

Subnet is used to extend the natural mask of some bit s to create the address of sub network ID.

Subnet ID: IP contains two parts, network ID and host ID. They are represented in class A, B and C.  For sub netting host ID is divided into two parts. Subnet ID and host ID.

Custom sub netting mask: to meet the network requirement a point is divided between network ID and host ID.   It decides how many sub-net are to be used and trading off bit allocation. It is better because there is no need to request new IP address ad it more flexible.

VLSM: variable length subnet mask. It is used to allocate IP address to sub nets according to their requirements.  It improves the flexibility of sub netting. 

 Download Technical Seminar Topic on Sub Netting and Super Netting PPT .

Sub Netting and Super Netting Differences:

Why there is need to scale the network. There are two issues for that.  Proper utilization of addresses used and proper scaling of the routing. These are major problem. They cannot be solved simply by installing router memory and increasing the size of routing tables.

It is a 32 bit binary number represented in 4 decimal values each represent 8 bits range from 0 to 255. It holds the address of network id and host id. It is also called internet protocol address.

Class A begins with 0xxx. It has value in range 1 to 126. Class B begins 10xx. It has range from 128 to 191. Class C begins with 110x. It has range from 192 to 239. Class D begins with 1110. It has range from 224 to 239. Class E begins from 1111. It has range from 240 to 254. In which CLASS A, B and C are most widely used. Class d is used for addresses and class e is used for future reference.

Sub Netting and Super Netting in TCS IP:

Subnet: IP contains two parts, network ID and host ID. They are represented in class A, B and C.  For sub netting host ID is divided into two parts.  Subnet ID and host ID.

CIDR: classless inter-domain routing.  It is an important feature as it eliminated the concept of class and it also supports route aggregation. It allocates address space efficiently. It is notate like xxx.xxx.xxx.xxx/n, where n is number of leftmost bit. For example: 192.168.12.2/23 represents the two c class networks like 192.168.12.0 and 192.168.13.0. 1o/8 denotes address from 10.0.0.0 to 10.255.255.255.

Assuming portion of ISP address block:

–          A aggregates 8/24 into single: (200.25.16.0/21);

–          B aggregates 4/24 into single: (200.25.24.0/22);

–          C aggregates 2/24 into single: (200.25.28.0/23);

–          D aggregates 2/24 into single: (200.25.30.0/23);

ISP will use 256/24 in its allocation for internet with advertisement like 200.250.0.0/16.

 Download Technical Seminar Topic on Sub Netting and Super Netting PPT .

Protection and Security in Operating System:

Security is the major clause for any operating system.  Operating system must be able to protect the data the environment it is being used like office, home; building etc. the data stored in the system must be safe from loss. Operating system must have the feature of data encryption so that it cannot be used by other person without the permission of owner.

There are two level meant for security purpose:

–          Physical level; computer systems must be physically secured.

–          Human level: screen must be handled with care.

Authentication: an authentication program should be there so as to identify user with key or user knowledge or finger print, or signature. It depends on the system to identify the user. Passwords are also the nice options with account ID.  They are best and easy to use. There is rare chance that password will be same to different users.

But password has some disadvantages also. It can be accessed by just guessing. Now there many application that have capability to break passwords. Some people are successful in guessing passwords if they know user personally. They try all combinations of letters and alphabets.

One time passwords:  they are to prevent the improper use password exposure.  User has secret code known as PIN.

Now threats like Trojan horse and trap doors are introduced. They misuse their environment. Worms are also the major threat, as they copy themselves and slobber system performance.

Viruses: they are major problems for systems. They spread automatically in the systems. They are nothing but the program that keep on executing and damage the working of system. 

Windows NT: they are designed for security purpose. For login in NT there is security access that that token ID and security ID. 

 Download Security in Operating Systems PPT .

Introduction to  Shape Memory Alloys Seminar Topic:

They are the metals alloys that do not lose their shape. Means their shape cannot be changed. They have martensitic phase transformation. They have two unique properties: shape memory effect and pseudo-elasticity.

Martensitic transformation means deformation of lattice of crystals. Means their shape is not changed only their lattice transformation have taken place.

• Shape memory effect: that transformation that begins in the absence of stress at a certain temperature M^0s. These transformations will be continued until the temperature of the stress reaches to M^0f. the reverse transformation will take place at A^0s and finally at this temperature A^0f material will be fully austenite.
• Pseudo-elasticity: this transformation takes place without the change of temperature. This process takes place upon loading and un-loading. It is just like the rubber effect.

Stress in an important factor in transformation. Because on applying stress temperature rises that may expand the metal.

Alloy composition controls the transformation from martensitic to Austenite form, as it controls the temperature.

SMA: first of all wire taken either hot or cold when they are martensitic form. Then this wire is loaded to heat till it reaches the temperature A^f. Then this wire is allowed to cool to reach the state of martensitic. This results in the formation of two stable shapes of solid.

shape memory alloys applications

They are used in daily life applications. They are widely used in bone treatment as plates. They are also used in dental and orthodontic application for band support during wiring process in teeth’s.  They are used as stent. They are bio compatible and have very large field of applications. It has very nice and reliable mechanical property. But it is costly too and one major problem is that it has poor fatigue property.

Download Technical Paper Presentation Topics For CSE  on Shape Memory Alloys.

Introduction to  CSE Neural Network and Optimization:

Neuro-computing is the advanced computing. It is just like the brain-like computing. It consist of large number of neurons that work simultaneously to solve a particular problem. Here neurons mean processing elements.

In terms of mathematic ANN is an complex non-linear function that adjust the parameters in such a way that ANN output is similar to measured output of an ANN. The first neural network formed was MADALINE, which was applied to the real world.

They are used to extract the pattern ad detect problems that are not possible for humans it solve them. It creates its own organization.  It is used for real time operations that are carried out in parallel. There are some artificial neurons created that are simpler than biological neurons.

Learning process: they are categorized in two general paradigms: associative mapping and regularity detection.  They are classified in two neural networks: fixed networks and adaptive networks.

Applications: it is very widely used in the real world. It used in the robotics systems, for image processing and speech recognition, military services and in health services also. 

They are divided into five categories: prediction, classification, data association, data conceptualization, and data filtering.

Back propagation algorithm: it start with random set of weight and values. One of the set is applied to input layer and then passed to output layers.

Call processing: it takes place when we dial call to any user from our handset or phone. Then a switch is established between origin and destination.

Switching: it determines the path of channel and grants a path to the original telephone we he call to someone.

Traffic control: it uses three simple layer that forward network with propagation algorithm.

 Download Paper Presentation Topics for CSE Neural Network and Optimization .

Introduction to WAN Design with Frame Relay:

It is generally used for wide area network. It operates in two layers in the OSI model. It is mostly used by telecommunication network. It gives the speed from 56K (DS0) to 43Mbps (DS3) depending on the service provider. There are different terms used in the frame delay like :

• CIR (committed information rate)
• DCE (data circuit equipment)
• DTE (data terminal equipment)
• DLCI (data link connection identifier)
• HDLC (high level data link  control protocol)
• LMI (line management interface)
• PPP ( point-to-point protocol)
• PVC (private virtual circuit)

First we have to purchase the service from frame delay provider. After purchasing we have to attach the point-to-point links (DTE to DCE) from available customer’s sites to cloud frame sites. Now PVC network has to be setup between the customer’s sites. These PVC endpoints can be identified with a data link connection identifier. Monitoring is done by using line management interface.

It is much cheaper than point-to-point connections for longer distances.  If congestion on providers’ network takes place then this can destroy the throughput. It is also difficult to ensure the quality of service as compared to other protocols.

Mitigating:It used to determine our present needs and plans that will be best for us and that make a growth. It used to check the service agreements. Contracts given by the service provider is also checked by it. It also has the decision making capability, based on price and needs.

Redundant links: they are alternative that do not rely on Schaumburg site.  They also examine business requirements to determine the best setup.

When these networks are designed correctly they can very cost effective network. This network is also very flexible.

 Download Technical Seminar Presentation Topics WAN Design with Frame Relay .

Introduction to  Pattern Matching Algorithm to Human Genome:

This technique is used to identify the same pattern from the large databases, without any use of costly software or hardware. It can be done by using mathematical calculations that are formulated into the algorithms that can be used for the pattern matching technique. This technique is widely used in the DNA matching.

DNA is divided into two groves, major grove and minor grove. They are aligned in two ways, pair wise alignment and multiple sequence alignment. There are different techniques used for pattern matching like: dot matrix, substitution, dynamic programming, and words method.

Pair wise sequence alignment: they can be written in 3 combinations like:

(A) ATTCGGCATTCAGTGCTAGA, (B) ATTCGGCATTCAGTGCTAGA, (C) ATTCGGCATTCAGTGCTAGA

(D) ATTCGGCATTGCTAGA, (E) ATTCGGCATTCAGTGCTAGA, (F) ATTCGGCATT – – – – GCTAGA.

Multiple sequence alignment: (J) TCAGAGCGAGA, (K) ATCCGGCCGGCAGCGAGA

(L) CAAAATTCAGAGCGAGA, (M)ATCCGCAGAGCCCGGGGAGA, (N) CCCGGCAGCGAGA

(O) ATCCGTTTTTTTTTGAGA.

They generally use dot matrix technique for pattern matching. When the scan the pattern several green, pinks, and red dots are formed which indicate their values like: red-A-A, green-T-T, pink-C-C, and blue-G-G.

Words method: it uses two techniques;

–          FASTA: first when the query is made it search the query from the hash table, if it founds then it align and score the matching segment.

–          BLAST:  it find each word from the query sequence by finding the list score a least T when score is using the pair score matrix.

Conclusion: these techniques are robust as they can be adopted to support current user and the challenges in computational world. It uses traditional methods like dot matrix analysis and some genetic algorithms. 

 Download Technical Seminar Topic on Pattern Matching Algorithm to Human Genome .