Factors of Network Performance

Total End-to-End Delay: The total time a packet spends in the network is calculated as the sum of four primary components: $d_{total} = d_{proc} + d_{queue} + d_{trans} + d_{prop}$

Processing Delay ($d_{proc}$): The time routers take to examine the packet header and determine where to direct the packet. This is typically microseconds and depends on the hardware efficiency.

Queuing Delay ($d_{queue}$): The time a packet waits in a buffer (queue) before it can be transmitted onto the link. This varies significantly based on the level of network congestion.

Transmission Delay ($d_{trans}$): The time required to push all the packet's bits into the wire. It is calculated as $L/R$, where $L$ is the packet length in bits and $R$ is the transmission rate (bandwidth) in bps.

Propagation Delay ($d_{prop}$): The time it takes for a single bit to travel from one end of the link to the other at the speed of light in the medium. It is calculated as $d/s$, where $d$ is distance and $s$ is propagation speed.

Traffic Shaping: This technique controls the volume of traffic being sent into a network in a specified period (bandwidth throttling) to optimize or guarantee performance and improve latency.

Load Balancing: Distributing network traffic across multiple servers or paths to ensure no single resource is overwhelmed, which directly reduces queuing delay.

Data Compression: Reducing the size of the data ($L$) before transmission. Since $d_{trans} = L/R$, smaller packets result in lower transmission delay.

Caching: Storing copies of data closer to the end-user (e.g., Content Delivery Networks). This reduces the physical distance ($d$) and thus minimizes propagation delay.

Unit Awareness: Always check if the question provides bandwidth in bits (bps) or bytes (Bps). Remember that $1 \text{ Byte} = 8 \text{ bits}$. Calculations for transmission delay must use bits.

Bottleneck Identification: In a multi-link path, the end-to-end throughput is limited by the link with the lowest bandwidth. This is known as the 'bottleneck link'.

Propagation vs. Transmission: If the distance is short (e.g., a LAN), propagation delay is often negligible. If the bandwidth is very high, transmission delay becomes negligible. Identify which dominates based on the scenario.

Sanity Check: If calculating delay for a transcontinental link, propagation delay should be in the tens of milliseconds (limited by the speed of light). If your answer is in seconds or microseconds, re-evaluate your units.

The 'Speed' Fallacy: Users often equate bandwidth with 'speed'. However, a high-bandwidth connection can still feel 'slow' if the latency is high (e.g., satellite internet), as the initial request takes a long time to return.

Ignoring Overhead: Students often calculate throughput by simply looking at the file size and bandwidth. In reality, TCP/IP headers and acknowledgments consume a portion of the bandwidth, reducing effective throughput.

Zero Jitter Assumption: In theoretical problems, jitter is often ignored. However, in real-world networking, inconsistent queuing delays mean packets rarely arrive at perfectly regular intervals.