Universal solution towards a problem-free data transmission network

Scope of Adjacent Nodes

Data transmission through a network can be viewed as a combination of transfers between adjacent nodes. The impediments that can arise when data travels from one node to the next is what this section aims to expose and address. Any bottleneck to network traffic can be resolved by isolating the root cause to an adjacent pair of nodes. Therefore, this proposed solution will confine comprehensive operations of a network transmission to any pair of adjacent nodes.

Static Data and Latency

The first issue to address is a conflict between data discovery and data transfer. The Internet Protocol applies a dynamic process to identify devices (such as routers) before transfering data. The Address Resolution Protocol creates a bottleneck that warrants an alternative. Instead of a dynamic mapping that identifies data transfer recipients, a static configuration that occurs at the instance of physical connection would be efficient. The data transfer process must be separated from data procurement and discovery.

There are several feasible means of identifying a directly connected device. One method involves triggering a dynamic process within the firmware of a device. This process would retrieve and store information such as the network ID, MAC address, IP address and Maximum Transmission Unit (MTU) of another device at the instance of a connection. This method is relevant to both wired and wireless technologies. The purpose of this retrieval process is to store device information that can facilitate a data transfer to the next node.

Static Data and Latency

The storage would occur mutually on each connected device. The stored network ID becomes useful when multiple devices are connected to a node (such as a switch) that branches to different networks. The firmware of each device provides information to the other connected device. The source of this information holds both fixed data that cannot change and variable data that may change. During data transmission, each device uses stored information from an adjacent device to forward data to the receiving link or endpoint.

Any discovery logic that would occur at data transfer should be performed by the firmware process during physical connection. The data transfer process should only apply static data that facilitates low latency. This technique not only minimizes transmission bottlenecks, but minimizes the chances for a malicious breach. A breach can be avoided for both wired and wireless cases since the target pair of devices are identifiable during a connection phase. This provides an opportunity to isolate them from foreign operators of any kind. The details of this implementation will be provided in another section.

Data Congestion

The second issue to address is data congestion. An approach that prevents congestion is crucial. Therefore, factors that can lead to congestion must be avoided. The possibility of factors that cannot be avoided (or can slip avoidance) must be taken into account. The approach suggested by this site is to prevent congestion at nodes that exclude endpoints. The endpoints must include a buffer that accomodates any data they produce. Therefore, data transmission should occur only within the capacity and capability of adjacent nodes.

The performance at each node should be allowed at optimum capacity. This means that adjacent nodes should exchange adequate information before a data packet is advanced and received. A transfer of packets should not be indiscriminate since the health and capacity of a node determines progress of data toward its destination at an endpoint. The key to such an approach is to facilitate a transfer activity that is predictable. An important step toward predictable activity is to eliminate factors that can foster ambiguity.

Data Congestion

The number of devices involved in a network transmission contributes to a complex environment that fosters ambiguity. A tolerance for arbitrary implementations and flexible standards intensifies this complexity. A coherence between logical operations of TCP/IP and the physical operations of devices become less feasible in such a scenario. The TCP and IP layers isolate source-to-endpoint activity from link nodes. Therefore, a mechanism is needed that allows any network to perform as a unified conduit of data transmission.

A critical approach would be to unify both layers by dividing the scope of data transfer completion into adjacent pair of nodes. The first step in acheving this goal is to apply static data that was obtained during a physical connection (as previously described). This step allows a transfer node to be in harmony with a recipient node during data transmission.

Data Corruption

The third important issue is data corruption. The conventional TCP/IP response to data corruption is to drop packets then expect retransmission by a source endpoint. The corruption of data in transit between endpoints may occur either at a source or participating nodes. Besides, the wired or wireless medium that connects devices must also be considered. A situation where data may be corrupted at multiple points (source, node and connecting medium) cannot be ignored. A further complication is that data corruption may originate from external factors interacting with points of a network.

The root cause of data corruption need not be identified during transmission. A more effective approach is to identify the first point affected by a particular corruption in a given direction of data flow then make a correction. The kind of correction taken would depend on either of two factors. These are the transient and persistent states of point disruption that can lead to data corruption. The appropriate correction for a transient kind of corruption is to resend data.

Data Corruption

However, retransmission at a source endpoint that is not the first point affected by data corruption is not optimal. Instead, a node adjacent to this first point should be the priority for retransmission. The factors that make this a feasible approach will be discussed later. On a further note, a persistent kind of corruption is best corrected by repairing or replacing the affected point (or resolving an external root cause). The first point affected by data corruption must be identified through a reliable means. One technique is to compare dynamic values that represent transient and persistent states with stored values.

An example of state validation is reflected in checksum calculations of the IP layer. A corrupted data packet would be dropped by a node due to failed checksum validation. This leads to a derived throughput difference between input and output data within the node. Such information is useful in measuring transient and persistent states of disruption during data tranfer between nodes. The key to such throughput validation begins with mapping granular components of a device to parameters that can be stored as reference values.

Data Corruption

This process would further involve a means of detecting parameters when a device is in an active state of operation. Such granular control and parameter mapping can become the basis of an enhanced IP layer that verifies data integrity, checks node health and provides a response to transient states within a cohesive physical component. These elements will be clarified in later sections.

This section has so far introduced elements of a proposed TCP/IP implementation. An elaboration of these elements relevant to optimizing and controlling data transfer between endpoints follows.

Previous Next