Network Standards Committee

Network Standards:
Networking standards can be classified as
* Proprietary
* Open
* De Facto

Most companies were run by skilled inventors, who came up with great ideas for new technologies and weren’t particularly interested in sharing them. This standard owned by one company or person. These standards are called proprietary standards.

If many companies get together and agree to cooperate, they can create an open standard instead of a bunch of proprietary ones. These standards are called open standards. These groups also work with standards approval bodies like the ITU and ISO to gain acceptance for their standards

De Facto is Latin for “in fact”. It is universal standard. There are 6 main organizations that maintain network standard internationally. They are

* ISO (International Organization for Standardization): Created in 1947. Note that, the short name is ISO not IOS. Organization chooses this name from Greek word “isos” means, “equal”. For that many people thinks that ISO-International Standard. But this is incorrect.

* IEEE (Institute of Electrical and Electronic Engineering): Manages MAC address for Network Interface Card. IEEE’s main claim to fame in the networking industry is the IEEE 802 Project. IEEE 802.15 defines wireless technology. IEEE 802.16 defines WMAN standard.

* EIA (Electronic Industries Alliance) & TIA (Telecommunications Industry Association): EIA/TIA 568 standard builds for Media installation and termination

* ANSI (American National Standards Institute)

* ITU (International Telecommunications Union): By the early 1970s, a number of countries were defining national standards for telecommunications. On March 1, 1993 the name of this committee was changed to the ITU-T (International Telecommunications Union-Telecommunications Standards Sector)

Internet Standards Organization:There are six organizations that are responsible for the development of the Internet’s architecture, standards and policies and related activities. These are

* ISOC (Internet SOCiety)
* IAB (Internet Architecture Board)
* IETF (Internet Engineering Task Force)
* IESG (Internet Engineering Steering Group)
* IRTF (Internet Research Task Force)
* IRSG (Internet Research Steering Group)

Among all, ISO is the most powerful organization. All are International standards except ANSI. ANSI standards only use in USA. Internet standards are developed by IETF. The IETF standards documents are called RFC (requests for comments)



Click image to go:

to share this page with your friends, select below

Selective Repeat Protocol

Selective Repeat Protocol:

Selective Repeat is part of the automatic repeat-request (ARQ). With selective repeat, the sender sends a number of frames specified by a window size even without the need to wait for individual ACK from the receiver as in Go-Back-N ARQ. The receiver may selectively reject a single frame, which may be retransmitted alone; this contrasts with other forms of ARQ, which must send every frame from that point again. The receiver accepts out-of-order frames and buffers them. The sender individually retransmits frames that have timed out.

Step 1:

This is almost similar to GBN but main advantage is receiver can store packets in its buffer

If sender sends packet 0,1,2,3 and packet 2 is lost.

Receiver will get packet 0,1,3. It will store all packets and ACK for 0,1,3 and it also shift its window from packet 2.

When sender get ACK of packet 0,1,3 it will shift its window from 2 3 4 5. But packet 2 has not ACK so sender will wait for time and send again 2 3 4 5.

When receiver will get packet 2 3 4 5 it will send ACK but found duplicate of 3 it will discard and shift window from 6 7 8 9

Step 2:

If total packet number window size 4 and packet number 0 1 2 3 then it can occurs exchange packet 0 of 1^st window with packet 0 of 2^nd window. That’s why packet number should be double than window size. Example window size 4 then packet number should be 0 1 2 3 4 5 6 7.

This is the main theme of selective repeat.

Now we can send error free data from PC1 to PC2.
But we use TCP/IP Protocol Suits. And TCP/IP Protocol Suits uses two protocols in Transport layer - TCP (Transmission Control Protocol) and UDP (User Datagram Protocol).

TCP use its own techniques to send error free data. So lets learn a little about TCP. TCP is a connection oriented protocol. What is connection oriented protocol lets see.


to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below

Go Back N Protocol

'Go-Back-N' Protocol:

The main theme of this protocol, sender sends multiple packets (Such As 1, 2, 3, 4, 5).

1

2

3

4

5

6

7

8

9

10

Receiver will give ACK for every packet sequentially. ACK1, ACK2, ACK3, ACK4, ACK5. Then sender will slide its window another 5 packets and send packets 6, 7, 8, 9, 10

5

6

7

8

9

10

11

12

13

14

In these ways GBN slide window, so GBN itself referred as a Sliding-window protocol.

Now we will see how sender and receiver send and receive data

Problem Studies 1:

If receiver gets 6, 7, 8, 10 packets and misses packet 9 then receiver will send ACK6, ACK7, ACK8

Then sender will slide its window another 5 packets and send packets 11, 12, 13 because sender knows he sent 9 and 10

8

9

10

11

12

13

14

15

16

17

But receiver end will discard packet 11, 12, 13 and will send no ACK. Sender will wait for a time and then send again packets 9, 10, 11, 12, 14. Receiver will send ACK sequentially for all packets.

Sender will slide its window from 14- 19 and will send 5 packets.

Problem Studies 2:

If receiver gets 6, 7, 8, 9, 10 packets then receiver will send ACK6, ACK7, ACK8, ACK9, ACK10

But ACK9 lost. Then sender will slide its window another 5 packets and send packets 9, 10, 11, 12, 13

8

9

10

11

12

13

14

15

16

17

There are some problem in GBN that if 1-100 packet send and packet 3 missing in this protocol sender have to retransmit again 3-100 packets. That is a wasting time. Cause receiver has no buffer to store. If there is buffer receiver can store 4-100 data and sender then has to retransmit only packet 3.

So there is another protocol which is doing this. This protocol is Selective Repeat Protocol.

to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below

Stop N Wait Protocol

Stop & Wait protocol

Step-1:
Sender:
1. Sender wait for the data from upper layer.
2. When sender gets data then it makes Data into Packet
3. But sender has no reliable channel to send data reliably. So it send unreliable data

Receiver:
1. When receiver get data from lower level, it think it is reliable packet.
2. Then it will extract the packets into data
3. Then it deliver data to its higher level

In this step had no solution. Cause receiver can get corrupted data. But how receiver will detect the data is corrupted. So to send error free data sender and receiver uses Error detection techniques. Sender calculates data and also sends the result and receiver calculates data and matches the result with sender’s value. There are many techniques to detect error.

Step-2:
Using one of these Error Detection techniques, sender will send data with result but when receiver will detect error or not, sender should know that. So receiver will send an acknowledgement (ACK) for error free data and receiver will send a negative acknowledgement (NAK) for corrupted data.

Sender:
1. Sender wait for the data from upper layer.
2. When sender gets data then it Checksum data
3. Then makes data into packets
3. Send unreliable data.
4. Then the sender wait for either ACK or NAK from the receiver.
If receiver get corrupted then it will send an NAK
5. The sender will send the packet again.

or If receiver get the error free data then it will send ACK
5. Then sender will again wait for the data from upper layer.

Receiver:
Receiver can get corrupted or uncorrupted data. If receiver get uncorrupted data from lower level then
1. It thinks it is reliable packet. Then it will extract the packet into data
2. Then it deliver data to its higher level
3. Then it make packet for send acknowledgement
4. Then it send acknowledgement to sender

On the other hand if it get corrupted data then
5. It make packet for a NAK and send it to sender

Problem:
When receiver send ACK/NAK, then ACK/NAK can be corrupted. When sender get corrupted ACK/NAK then it send again the packet. So at the receiver end data can make duplicate. Example:
we want to send --- today is our networking class. 
Receiver get --- today is our our networking class class

Now what can we do to solve this problem? We can solve this problem by giving packet number. We can give packet number like this-1, 2, 3, …1000...and more. But it is a problem if the number of packets are huge then number will be big problem. So we can solve this by 0 and 1. Cause duplicate occurs in two packets. So consider this numbering all packet by 0 and 1.

Step 3:
Sender:
1. Sender wait for the data from upper layer.
2. When sender gets data then it Checksum data
3. Then makes data into packets
3. Sends unreliable packet 0
4. Then the sender wait for either ACK or NAK from the receiver.
If receiver get corrupted packet 0 then it will send an NAK, The sender will send the packet 0 again.

or If receiver get the error free data then it will send ACK
7. Then sender will again wait for the data from upper layer and send packet 1

Receiver:
When it needs packet 0, it can receive 3 types of data - Corrupted packet 0, uncorrupted packet 0 and uncorrupted packet 1.
If receiver get corrupted packet 0
1. It will send a NAK to sender

Or when receiver get uncorrupted packet 1.
1. Then it should be send NAK but receiver yet has no power, so it sends ACK

Or when receiver get uncorrupted packet 0
1. Then it will extract the packet into data
2. Then it deliver data to its higher level
3. Then it make packet for send acknowledgement
4. Then it send acknowledgement to sender

Problem:
Here we saw the duplicate ACK. So we can number also ACK and omit the NAK.

Step 4:
Sender:
1. Sender wait for the data from upper layer.
2. When sender gets data then it Checksum data
3. Then makes data into packets
3. But sender has no reliable channel to send data reliably. So it send unreliable data
4. Then the sender wait for either ACK either packet 0 or 1.
If sender get corrupted ACK or ACK for packet 1, The sender will send the packet again.

or If sender get the uncorrupted ACK or ACK for packet 0, then sender will again wait for the data from upper layer

Receiver:
When it needs packet 0, it can receive Uncorrupted packet 0 and uncorrupted packet 1.
If receiver get uncorrupted packet 1
1. Then it send ACK1

Or when receiver get uncorrupted packet 0
1. Then it will extract the packet into data
2. Then it deliver data to its higher level
3. Then it make packet for send ACK 0
4. Then it send ACK 0 to sender

Problem:
Data can be lost on the way. ACK can be lost. What will do sender & receiver. How many times they will wait. They will wait for a certain time. So we have to use timer in this program. What will we do with timer?
1. start timer each time when packet sent
2. respond a timer interrupt
3. stop timer

Step 5:
Sender:
1. Sender wait for the data from upper layer.
2. When sender gets data then it Checksum data
3. Then makes data into packets
3. Sender sends unreliable data
4. Then the sender wait for ACK either packet 0.

If corrupted ACK for packet 0 or ACK for packet 1
5. Then it will do nothing. It will wait for time out. If Timeout -
6. Then it will send packet 0 again
7. Start the timer

Or if get uncorrupted and ACK 0
8. Then sender will do nothing. It will wait.
During waiting one thing can be happen receiver sends ACK but didn’t reach in time. Sender will count time out and send again. It will make duplicate at receiver. Receiver also sends again that data reached.
8. Then sender will do nothing

Receiver:
1. When it needs packet 0 and it will get uncorrupted packet 1
2. Then it will do nothing. It will wait for time out. If Time out
3. Then it will send packet again

If receiver get uncorrupted ACK 0
1. Then it will extract the packet into data
2. Then it deliver data to its higher level
3. Then it make packet for send ACK 0
4. Then it send ACK 0 to sender

As a sender same thing can happen at receiver. Sender sends data but will not reach in time. So at receiver end wait for time out and after timeout it will send to send again. Sender will send again but by this time 1st data has reached so second data is duplicate.
Then receiver will do nothing.
This is the main theme of Stop & Wait Protocol.

To follow this protocol our 3rd problem that was error detection is solved. Stop and wait protocol is long time process handle with only one packet at time. As a result it takes time. If there any other protocol who can handle multiple packets at a time so it will faster than before.
There is another protocol Go-Back-N.

to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below

Data into Small Packets

Data Packets:

We need to send messages from one end to another. A big file cannot be transmit. So we need to divide data into packets of fixed or variable size. The size of the packet is determined by the network and the governing protocol.

Depending on Layer and the type of network, packets may be referred to by another name 
* Segment (Transport Layer) 
* Datagram (Network Layer) 
* Frame (Datalink Layer)

Most packets are split into three parts:

1. Header - The header contains information. Such as source and destination address.
2. Payload - Also called the body or data of a packet. This is the actual data that the packet is delivering to the destination.
3. Trailer - The trailer, sometimes called the footer, typically contains a couple of bits that tell the receiving device that it has reached the end of the packet.

For an example, you send an email to a friend. The e-mail is about 3,018 Byte in size. According to IMP in 1969 (RFC7), the maximum massage length 8080 bits or 1010 Bytes. In this massage header and marking is 32 bits or 4 Byte. So data should be 1006 Bytes. Then it will make 3018 Byte mail into 3 packets.

This problem was solved. We can send data as packets. But really data reached at destination properly? Data can be corrupted or data can be lost on the way. So receiver may not get actual data. This is our 3^rd problem. For an Example

I sent - I have a dog

He/She received
One packet corrupted - I have a dgo
Or packet Lost - I a dog

This job is also done in transport layer. So all began to make the protocol that provides reliable data transfer. Few protocols are

* Stop & Wait
* Go back N
* Selective Repeat

to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below

OSI Model

Layers:
In 1974 ISO (International Organization of Standard) start to make a model of layers. In 1977 they finalized this model and named OSI (Open System Interconnection) Model.

OSI Model:
There are 7 layers in OSI Model.
7. Application Layer
6. Presentation Layer
5. Session Layer
4. Transport Layer
3. Network Layer
2. Data link Layer
1. Physical Layer

The work is start from Application layer. But the Counting of Layer is start from Physical layer. So Physical Layer is Layer-1.

To remember all layer easily, personally I remember APSeTNeDaP.
We know that different layer responsible for different job when we send data from one PC to another. Now we will see shortly what the certain job for each layer is. Later we will know details.

The Application layer is responsible for providing services to the user

The Presentation layer is responsible for translation, compression and encryption

The Session layer is responsible for establish, manage and terminate sessions.

The Transport layer is responsible for the delivery of a message from one process to another.

The Network layer is responsible for the delivery of individual packets from the source host to the destination host.

The Data link layer is responsible what it done in NIC such as modulation demodulation.

The Physical layer carries the data from one NIC to another NIC blindly.

If we compare what we have done with OSI Model, we found 3 layers. the layer with we work for sending data that is Application Layer, any job at NIC card is Data Link Layer and sends data through media in Physical Layer.

Different layer uses different protocols. The protocols of the various layers are together is called Protocol Stack. The Internet Protocol Stack consists of 4 Layers implemented from OSI Model. This are-

1. Process/Application Layer (combination of Application + Presentation + Session Layers of OSI Model)

2. Host to Host Layer (Transport)

3. Internet Layer (Network)

4. Network Access Layer (Data link + Physical)

TCP/IP protocol suite was generally designed to provide implementation of the networking stack from the Network Layer / Internet Layer (layer 3) and above. In most classical networks, TCP/IP doesn’t define any protocols operating at Network Access Layer (Data link + Physical). TCP/IP assumes that layer two functionally is provided by a WAN or LAN technology like Ethernet, Token Ring or IEEE 802.11. Among these technologies Ethernet is by far the dominant technology. We will learn later about this technology.

We have connected PC with cable, connector and NIC. Now I want to transfer data to my friend’s PC but my PC doesn’t know where to send. 
To find a house we need the house number, road number and area name. In this way PC also need numbering to find destination. What we can do? Numbering two PCs that both PCs can recognizes each other with that number. This number is known as IP address. For applying IP address there must be a rule, that means a protocol. The rules for IP address are defined in Internet Protocol (IP). IP Protocol works at Layer 3 (Network / Internet Access Layer). Later we will know details about IP Address and Internet Protocol. Now we will give number (IP) for PC are 192.168.1.1 and 192.158.1.2. And there is an option of subnet mask we will give each PC 255.255.255.0
Now each PC got an IP address by using Internet Protocol and data will send with source and destination address.

Source Address	Destination Address
Data

We solved 1^st problem of modulation and demodulation. But 2^nd problem arises when we send a big file across to the other PC. Cause Ethernet allows transferring data only 1500 bytes at a time. What can we do?

We can do one thing - divided data into pieces. Each pieces are called in networking world is Frame or packet. How we can make packets. So lets see how data makes packet or frame.

to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below

Protocol

What is Protocol? 
Protocol is a set of rules that computers follow to talk to each other. It means computer share data which must have some rules and all computer must have to follow these rules.

There are so many protocols created for running the network smoothly. The certain protocols make a package called Protocol Suit. There are many types of protocol suits.

1. TCP / IP (Transmission Control Protocol / Internet protocol):
ARPA (Advanced Research Projects Agency) developed their research network called ARPAnet was designed to use a number of protocols. This protocol bundle was called TCP (Transmission Control Program) used in their existing technologies and written in 1973 published in RFC 675, December 1974. Testing and development of TCP continued for several years. In March 1977, version 2 of TCP was documented. Jon Postel (one of the most important pioneers of the internet and TCP/IP) splitting of TCP at the transport layer and IP at the network layer; thus the name TCP/IP given and version 3 begun and written in1978. January 1, 1983 TCP/IP officially declared as new standard protocol for ARPAnet instead of NCP protocol. TCP/IP is the most popular protocol suits. Because it can supports any kind of operating systems.

2. IPX / SPX (Internetwork Packet eXchange / Sequenced Packet eXchange):
For the Novell NetWare operating systems

3. AppleTalk:
For Apple Macintosh computer  

We will use TCP/IP protocol suite in our network. There are many protocols in TCP/IP protocol suite. To install protocols we must install NIC. Protocols are working step by step.

So question is, which protocol works first and which works later among so many protocols? So full network is logically divided into several steps and certain protocols are work for each step.
For example, we found the three steps in what we have done. One step is we sending data, 2^nd step handles conversion in NIC, prepares data to send and 3^rd step is sending data via media.
Actually steps are called Layer. Different layer works certain different job. Why need these layers? For example, how we eat, by hand or by leg? We know by hand. So a specific portion in our body do this work. Another example, we know the importance of making departments in office. Such as Accounting Dept works only account related works. Admin Dept looks after administration of office. In network, when we send something, PC does the various types of jobs. Each job done by different protocols. So logically it divided some layers that the network will be easy to understand. 
Now we will learn a little about these layers which is called OSI Model in the network field.

to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below

NIC and Connectors

Network Interface Card (NIC):

To connect with cable, a computer uses a Network Interface Card (NIC)

Each LAN Card has an ID number, which is unique number in the world. This id is called MAC address (Media Access Control). IEEE manages the MAC address space. When Company wants to manufacture adapters; it purchases a chunk of the address space for a nominal fee. 
Suppose a MAC address is 12-34-56-78-90-AB
Six Hexadecimal Number. 1st 3 Hexadecimal numbers are given from IEEE and are known as the Organizationally Unique Identifier (OUI). Last 3 Hexadecimal numbers are given by manufacturer.

(Now we have got cable and LAN card but how we connect card and cable. We need Connector. There are different connectors for different cable.) 

Connectors:
Considering connector or cable, LAN port should be compatible. Connectors for Cables are -

Coaxial Cable:
Thin net – RG-58
Thick net - RG11, RG8 Most use coaxial cable is RG-59 category and connector we use for this cable is BNC

Lets see how to connect connector with coaxial cable
1. Cut the Outer Black Jacket cable half inch

2. Carefully pull back the Shield outside the outer cover

3. Trim back the plastic (usually white) from the inner core cable

4. Push the connector down over the cable end so that the copper core of the coaxial cable sticks out

5. Screw the connector into the end of the cable

Twisted Pair Cable:
For Twisted Pair Cable, we use RJ45connector. RJ means Registered Jack.
There are 8 pins of connector.
1. TX+
2. TX-
3. RX+
4. Unused
5. Unused
6. RX-
7. Unused
8. Unused

7&8 for future use such as PoE but 4&5 always off. that's why its called RJ45.
The positive and negative connectors in the cabling mentioned, form a balanced pair for signals. They typically carry identical signals of opposite polarity, i.e. a HIGH bit may be +5 Volts on the + conductor, and -5 Volts on the - conductor. The corresponding LOW bit would then be -5 Volts on the + conductor, and +5 Volts on the - conductor.
Having such signals of opposite polarity in a twisted pair of wires helps in reducing susceptibility to common-mode noise / EMI. Because the voltage difference between the two conductors for any valid bit is twice the actual signaling voltage, this has the effect of doubling the strength and hence the detection of the signal at the receiving end.
What will be the color combination of cable? Two organizations fixed an international standard for this cabling EIA & TIA. So if we do something should try to follow International rules. EIA/TIA 568 standard builds for Media installation and termination.

EIA/TIA 568A defined color-coding for RJ45 connector is-
1. Green-White
2. Green
3. Orange-White
4. Blue
5. Blue-White
6. Orange
7. Brown-White
8. Brown

EIA/TIA 568B defined color-coding for RJ45 connector is-
1. Orange-White
2. Orange
3. Green-White
4. Blue
5. Blue-White
6. Green
7. Brown-White
8. Brown

Note that it is very important that a single pair be used for pins 3 and 6. If one conductor from one pair is used for pin 3 and a conductor from another pair is used for pin 6, If this is not done the receive data will be split between 2 pairs and performance will degrade.
We have learned cabling in two ways - 568A and 568B. If we match color coding both end of the cable 568A or 568B this is called Straight Through Cable. If one side is 568A and another side is 568B this is called Cross Over Cable.
Now question is that when we will use cross over or straight through cable? If the devices at the both ends are same then we will use cross over cable and if the devices at the both ends are different then we will use straight through cable. (There is another definition we will learn later.)
There is another type of cabling called Roll Over. One side is inverse of another side. Suppose if one side is 568A then another side will be Brown - Brown White – Orange – Blue white – Blue – Orange white – Green – Green white. Roll over cable used for configuration purpose. Suppose any router, manageable switch etc. To connect switch/router it needs to attach a console connector in one side for the console port. That’s why it is sometimes called console cable.
To punch cable with RJ45 Connector, we need a tools called Crimper.

Fiber Optic Connectors:
There are different types of Fiber Connectors.

Most Used Connectors are SC (Subscriber Connector), FC (Fiber Connector), LC (Lucent Connector). But we have no available devices to connect connector with fiber cable. So we buy Fiber Patch Cord (built in connector with few meters fiber).

Patch Cord connector can be different types whatever you want. In the figure there is a SC/SC Patch cord. What will we do just cut the patch cord half and joint the fiber with 1 piece Patch cord by Splicing Machine. This is the easiest way done by all Internet Service Provider.

In this way we use different types of connector in fiber.

We have connected our two PCs with a cross over cable according to EIA/TIA rules via NIC (Network Interface Card).
But out 1st problem arises that these cables can pass only current voltage, not anything else. Then how we will send data?
Think carefully cable is either passing current or not. That means when switch on current passing through cable, when switch off, cable passing nothing. Using this on off activity we can send signal to receiver. If we let the current on is 1 and off is 0 then receiver will get combination of 0 and 1.

For an example (This is not actual value), PC1 wants to send "How are you" to PC2. If we let 
How are you = 1010
then PC1 will switch on and voltage starts to pass, receiver will get 1. Switch off voltage is not passing, receiver will get 0. Again switch on and voltage starts to pass, receiver will get 1. Switch off voltage is not passing, receiver will get 0. So receiver will get 1010 combination. So PC2 will understand, PC1 sent "How are you"

Isn't that nice? Now we have to convert all data into 0 and 1 anyhow but in a unique way so that 01 combination never match with one another. Then we can send data using current signal. 
That's why ASCII Code was created. We will know little about ASCII.

to continue reading click on Next

Back      Next

Click image to go:

to share this page with your friends, select below