sliding window protocol window size
There was no fixed window size in the sliding window protocol. It would be determined based on various factors such as the specific network conditions, the application scenario of the protocol, the size of the sending buffer, and the receiving ability of the receiver. It was mentioned in Flink's window architecture related content that, for example, the window size could be set to 5s (an example of a step size of 2s). This was only an example of a specific application (a windowing application under the Flink architecture). It could not represent the common window size settings of the sliding window protocol. The purpose of the sliding window protocol was to control the flow to avoid congestion. The window size was determined to ensure the reliable transmission of data and to increase the network throughout as much as possible. For example, in network data transmission, if the sending party has a strong sending ability and the receiving party has a limited receiving ability, the window size needs to be set reasonably according to the receiving party's receiving ability and other factors, so that the sending party can send multiple data packets at the same time without receiving an acknowledgment, without causing problems such as network congestion or data loss due to sending too many data packets exceeding the receiving party's ability. "Choose" was equally exciting. Everyone was welcome to read it!
sliding window protocol
The sliding window protocol was a full-Duplexed protocol that included a sending window and a receiving window. The transmit window is the number of frames that have been sent but have not been acknowledged, and the receive window is the number of frames that are expected to be received. For example, if the default window is 1 and the serial number length is 3 digits, there is a specific sending process. It was a flow control technique in the transmission layer of the network protocol. In the protocol, the sliding window mechanism was used for flow control. The receiver would adjust the window size according to its own reception ability. This window size indicated the amount of data that the receiver could currently receive. The sender would control the speed of sending data according to the window size fed back by the receiver, thus ensuring that the sender would not overflow the receiver's buffer due to sending too much data. " Choose " was equally exciting. Everyone was welcome to read it!
Continuous arq protocol and sliding window protocol
In the continuous ARQ protocol, the sender maintained a sending window of a certain size, and the packets within the window could be sent continuously without waiting for the other party's acknowledgment. The sender would slide the sending window forward by one packet every time it received an acknowledgment. The receiver usually uses the method of accumulating acknowledgement, that is, after receiving several packets, it sends an acknowledgement to the last packet that arrives in order, indicating that all the packets up to this packet have been received correctly. However, this method has shortcomings. For example, when the middle packet is lost, the sender cannot know the situation of the subsequent packets, and may need to re-transmit more packets. The sliding window protocol involves a sending window and a receiving window. The transmit window is the range of sequence numbers of the frames that have been sent but have not been acknowledged, and the receive window is the range of sequence numbers of the frames that are expected to be received. When the default window size is 1, you can operate according to the specific sending process. The sliding window protocol could improve the channel utilization and was a reliable transmission mechanism for data communication. The continuous ARQ protocol can be seen as a specific application form of the sliding window protocol under certain conditions (such as the cumulative acknowledgment method). They are all protocol mechanisms that exist to achieve effective and reliable data transmission in the network. "Choose" was equally exciting. Everyone was welcome to read it!
Is the select replay protocol a sliding window protocol?
Yes, the selective replay protocol was based on the sliding window flow control technology. It was a sliding window protocol. " Choose " was equally exciting. Everyone was welcome to read it!
Choose the size of the window for the replay protocol.
For the selective replay protocol, the window size is <= 2^(m - 1). When n bits are used to number the frames, in order to prevent the old and new receive windows from overlapping after the receiver moves the window forward, the maximum size of the transmit window does not exceed the range of the sequence number, that is, WT <2n-1. When the transmit window takes the maximum value, it is found to be the same as the transmit window and 2n-1 (in most cases, the transmit window is equal to the receive window and 2n-1, because this is the most efficient). If the value of the transmit window is greater than 2n-1, it may cause the new and old receive windows to overlap. "Choose" was equally exciting. Everyone was welcome to read it!
What is the window size for selecting the replay protocol?
For the selective replay protocol, the window size was <=2^(m - 1), and the transmit window could not be larger than the receive window. To maximize efficiency, in most cases, the transmit window was equal to the receive window and was equal to 2n - 1 (n was the number of bits used to number the frame). However, this value must not exceed half of the range of the serial number to avoid the overlap of the new and old receive windows at the receiver. "Choose" was equally exciting. Everyone was welcome to read it!
Choose the maximum window size for the replay protocol.
If n bits are used to number the frame, the maximum size of the transmission window of the selected replay protocol does not exceed half of the range of the sequence number, that is, WT <= 2^(n - 1). When the transmit window is the maximum, it is the same as the receive window. In most cases, the transmit window is equal to the receive window and is equal to 2^(n - 1), because the maximum efficiency can be achieved in this case. At the same time, in order to prevent the new receiving window from overlapping with the old receiving window after the receiving end moves the window forward, the size of the sending window should meet this limit. Otherwise, the new and old receiving windows may overlap. "Choose" was equally exciting. Everyone was welcome to read it!
Choose the size of the reception window of the Retransmission protocol.
By default, the size of the receive window is equal to the size of the send window in the selective replay protocol (SSR), but no information about the specific size was found. In the example where the frame sequence number is represented by 3 bits, the principle that the sending window meets is that the size of the sending window is equal to the receiving window, but the specific size of the receiving window is not specified. "Choose" was equally exciting. Everyone was welcome to read it!
Fallback n frame protocol sending window size
The size of the sending window of the fallback N-frame protocol is greater than 1, and multiple data packets can be sent consecutively. " Choose " was equally exciting. Everyone was welcome to read it!
Backward n frame protocol sending window size
When n bits are used to number the frames, the size of the transmission window should be less than 2n - 1. If the size of the transmission window is larger than 2n - 1, the receiver will not be able to distinguish between the new frame and the old frame. For example, 8 different serial numbers can be made with 3 bits, and the maximum number of transmission windows should be 8, but in fact, setting the transmission window to 8 will make the protocol unable to work under certain circumstances. " Choose " was equally exciting. Everyone was welcome to read it!
How to set the size of the window for selecting the replay protocol?
For the selective replay protocol, the window size <=2^(m - 1). Firstly, the sending window could not be larger than the receiving window, or else the receiving window might overflow. Secondly, the pipeline grouping of the sending window must be maximized, but it must be ensured that there was no ambiguity. Assuming that the maximum sequence number is 7, that is, 0, 1, 2, 3, 4, 5, 6, 7, and the size of the sending window is 5, when the sending window sends 0, 1, 2, 3, 4, assuming that the receiving window receives all of them, the receiving window will move forward 5 times, and the receiving window will expect to receive 5, 6, 7, 0, 1. If the sending window does not receive any ACK, the receiving window will think that the re-sent 0, 1, 2, 3, 4 are new packets. Because the sending window <= receiving window. To maximize the send window, send window = receive window. Assuming that the transmission window is m, the reception window is also m. When the sending window sends m packets, the receiving window moves forward m, and the receiving window is m + 1,m + 2,... 2m。To avoid ambiguity, 2m <= the maximum number of the sequence. "Choose" was equally exciting. Everyone was welcome to read it!