NET note-delay, packet loss and throughput in packet switching networks

Grouping starts from a single host (source), transmits through a series of routers, and ends its journey in another host (destination). When grouping from a node (host or router) along this path to a successor node (host or router), the packet undergoes several different types of latency at each node along the way. The most important of these delays are node processing delay (nodal processing delay), queue delay (queuing delay), transmission delay (transmission delay), and propagation delay (propagation delay), These time-delay totals add up to the total nodal delay of the node. Many internet applications, such as search, Web browsing, e-mail, maps, instant messaging, and voice over IP, have a high performance impact on network latency. In order to gain an in-depth understanding of packet switching and computer networks, we must understand the nature and importance of these delays.

Types of delay

Let's look at these delays in the environment below. As part of the end-to-end path between the source and destination, a packet is sent from the upstream node through router A to Router B. Our goal is to depict the node delay at router A. It is important to note that router A has an outbound link to Router B. The link is preceded by a queue (also known as a cache). When the packet arrives at router a from the upstream node, router a examines the header of the packet to determine the appropriate outbound link for that grouping and directs the grouping to that link. In this example, the link to the packet is the one that leads to Router B. The grouping can be transferred on this link only if no other groupings are in transit on the link and no other groupings are queued in front of that queue; if the link is currently busy or there are other groupings already enqueued on that link, the newly arrived group will participate in the queue.

Total node delay:

(1) processing delay (nodal processing delay)

The time required for the

to check the packet header and determine where to orient the group is part of the processing delay. Processing delays can also include other factors, such as the time required to check for bit-level errors, which occurs during the transfer of these packet bits from upstream nodes to router a. The processing time delay of a high-speed router is usually a microsecond or a lower order of magnitude. After such a node is processed, the router directs the packet to the queue before it leads to the router B link.
 
(2) queue delay (queueing delay)

In the queue, it undergoes a queued delay when the packet is waiting for transmission on the link. The time-to-queue extension of a particular grouping depends on the number of packets that are queued for transmission to the link in advance. If the queue is empty and no other groupings are currently being transferred, the queued delay for that grouping is 0. On the other hand, if the traffic is large and many other groupings are waiting for transmission, the queuing delay will be long. We will soon see that the number of packets arriving at the packet expectation is a function of the intensity and nature of the traffic reaching that queue. The actual queueing delay can be in milliseconds to microseconds of magnitude.  
(3) transmission delay (transmission delay) assumes that the packet is transferred first-to-first, which is common in the packet-switched network, and only when all the packets that have been reached are transferred, the packets just arrived can be transferred. The length of the packet is represented by the L-bit, and the link transfer rate from router A to Router B is represented by R bps (that is, b/s). For example, for a 10Mbps Ethernet link, the rate is r=10mbps, and for a 100Mbps Ethernet link, the rate is r=100mbps. Transmission delay is L/R. This is the time required to push (transmit) all the grouped bits to the link. The actual transmission delay is usually in milliseconds to microsecond magnitude.
 
(4) propagation delay (propagation delay) once a bit is pushed to the link, the bit needs to be propagated to Router B. The time required to propagate from the beginning of the link to Router B is the propagation delay. This bit is propagated at the propagation rate of the link. The propagation rate depends on the physical media of the link (i.e., fiber, twisted-pair copper wire, etc.), the rate range is 2x108~3x108m/s, which is equal to or slightly less than the speed of light. The propagation delay is equal to the distance between two routers divided by the propagation rate. That is, the propagation delay is d/s, where D is the distance between router A and router B, and S is the propagation rate of the link. Once the last bit of the packet is propagated to Node B, the bit and all the preceding bits are stored in router B. The entire process will continue as Router B performs the forwarding. In the WAN, the propagation delay is a millisecond magnitude.
 
(5) Comparison of transmission delay and propagation delay

It is sometimes difficult for novice computer users to understand the difference between transmission delay and propagation delay. The difference is subtle and important. Transmission latency is the amount of time the router will need to group out, which is a function of packet length and link transfer rate, regardless of the distance between the two routers. On the other hand, propagation delay is the time required for a bit to propagate from one router to another, which is a function of distance between two routers, regardless of packet length or link transfer rate.

queue delay and packet loss

Queue Delay

ratio LA/R is called flow intensity

If la/r >1, the average rate at which the bit arrives in the queue exceeds the rate at which the queue is transferred, the increase in the queue tends to be unbounded, and the queuing delay tends to be infinite. Therefore, there is a golden rule in traffic engineering: The flow intensity can not be greater than 1 when designing the system.

If la/r <=1, the nature of the arrival traffic affects the queueing delay. For example, a packet arrives periodically, that is, every L/R s arrives at a grouping, and each packet arrives in an empty queue, so there is no queuing delay. If the packet arrives in bursts rather than periodically, there may be a large average queueing delay, assuming that each L/R s reaches n groupings at the same time, the first packet has no queueing delay, and the nth packet has a queueing delay of (n-1) L/R S.

Drop Packets

The queuing capacity is limited, and when a packet arrives at the router, it is found that there is no local storage and the router discards the packet. The performance of a node is often measured not only by time-delay, but also by the probability of packet loss.

throughput Transfers a file over a computer network from host A to Host B, and any instantaneous throughput is the rate at which Host B receives the file. If the file consists of F-bits, and Host B receives all bits with a T-second, the average throughput of the file transfer is f/t bps.