blog posts

Some Networked Devices Improve The Channel By Compressing The Data They Send. Most Modems Use This Feature. 

Network Throughput: If The Size Of A File, For Example, 64 KB, Can Be Reduced by Compression, the time required for transfer is reduced.

This can be hidden from the user, so a compressed file can be transferred significantly faster than expected. Because this ‘hidden’ compression cannot be easily disabled, a file that cannot compress must be used to measure throughput by transferring files and scheduling transfer time. Pre-compressed files, such as zip files, are often used.

Assuming it is impossible to compress the data to transfer a 64 KB file over a 64 kbps channel, the theoretical time required is 8.192 seconds. This is the minimum transfer time, but it takes longer in practice. This is due to the overhead effect of formatting the data in the agreed state.

Therefore, both sides of the relationship consistently view the data.

Network Throughput

At least two issues with compressing files are not immediately apparent.

1. The throughput of the network that performs the compression does not improve with compression. From a bottom-up perspective (server to client), compression enhances throughput. This is because the information content increases with compression for the same amount of information transmitted.

2. Compressing files in client and server systems requires more processor resources. Currently, the server compresses the files using its processor, which is false. The client also decompresses the files upon receipt.
This can be considered one of the necessary costs (for server and client systems) to increase end-to-end throughput (even though throughput on the same network does not change).

What is operational capacity?

A network’s throughput can be measured using tools available on different operating systems. This page explains how these instruments are set up and the issues associated with these measurements.

When measuring network throughput, people often want to know the maximum data rate in a communication link or network access, expressed in bits per second.

A common way to measure this quantity is to transfer a large file from one system to another and measure the time required to complete the transfer, or to copy the file.

Operating power is then obtained by dividing the file size into megabits per second, kilobits per second, or bits per second.

Unfortunately, the results of such an experiment are in the range of good throughput, which is usually lower than the theoretical maximum throughput. This leads people to think that their communication link is not working correctly.

 In addition to transmission overheads, including downtime, TCP receiving window size, and system constraints, there are many overheads in goodput, so the calculated goodput does not reflect the maximum available throughput.

Typically, people use abbreviations that are often used interchangeably. For file sizes, they usually use ’64k ‘instead of ’64kilobytes’ or ‘100meg’ instead of ‘100megabytes’.

 When referring to bit rate, use the terms throughput, bandwidth, and speed to indicate the circuit speed as ’64k’ or ‘2meg’, meaning 64 kbps or two megabits per second. They give. However, a 64k circuit cannot transfer a 64k file in one second.

 This may not be obvious to people unfamiliar with telecommunications and computing, as they sometimes make mistakes.

Overheads and data formats

Most people use the serial link, which is also called the ‘asynchronous start-stop’ or ‘asynchronous’ link. If you use a modem externally connected to your home or office computer, your communication link may be established through an asynchronous serial connection.

This link’s advantage is that it is simple and can be implemented using only three wires: send, receive, and ground signal (standard signal).

 In the RS-232 protocol, bit ‘0’ has a negative voltage relative to ground, and bit ‘1’ has a positive voltage relative to ground. An inactive RS232 signal has a negative voltage.

So, any of the above descriptions about bits ‘0’ and ‘1’ are upside down here.

The leading bit of each byte has a high voltage, indicating that the next bit is the first bit of the serial data.

Asynchronous serial data at the TTL level ‘0’ at high voltage and ‘1’. A TTL/RS232 converter handles TTL-level data, usually a Maxim chip such as the MAX232, which also inverts and converts the data level.

All signals within a device are equipped with an RS232 connection and usually operate at internal TTL levels (0 and 5 volts, 0 and 3.3 volts, or 0 and 3 volts). Converting to RS232 is the last step for output data, or the first for input data.

In fact, more items, such as the one-bit transfer rate, the number of bits per character, the parity bit, and the number of stop bits (meaning the end of a character), are agreed upon to advance the transfer.

For example, the agreed feature 9600-8-E-2 means that the channel has a transfer rate of 9600 bits per second, with 8 bits per character, an even balance bit, and two stop bits.

These specifications in serial connections are usually set to 9600-8-N-1 (9600 bits per second, 8 bits per character, no parity bit, and one stop bit).

A total of 10 bits are transmitted to send each 8-bit character (one start bit, 8 data bits, and one stop bit), which means 25% overhead.

Therefore, an asynchronous serial link at 9600 bits per second (1200 bytes per second) does not transmit data at that rate.

This situation can get worse. If a balanced bit and two stop bits are used, the overhead of each 8-bit character will be 4 bits (one start bit, a balance bit, and 2twostop bits), equivalent to a 50% overhead.

In this case, a channel with a speed of 9600 bits per second can transmit 12,600 bytes per second (equivalent to 800 bytes per second).

Transfer speeds in asynchronous serial relationships are typically supported up to 230.4 kbps.

If the balance bit is not specified in the channel settings and a stop bit is used, the transfer rate will be 23.04 KB/s.

The advantage of the asynchronous serial connection is its simplicity. One disadvantage is its low transportation efficiency. We can overcome this by using a simultaneous interface.

In this type of interface, a clock signal is added via a separate wire, and the bits are transmitted according to the clock.

The interface no longer pays attention to each character’s start and stop bits; however, a mechanism is needed to ensure that the sending and receiving clocks are synchronous.

For this purpose, the data is divided into frames of different characters using known separators.

FAQ