This communications protocol governs the establishment and use of forced-entangled subspace transmission beams over multiple adjacent subspace domains. The protocol establishes a subspace lock onto a destination device and forms a unique quantum signature node within both the sender and receiver. Each node is then forced into an entangled state using a quantum inversion tunnel. This process establishes a unique connection between nodes through the subspace connection.
Initialization
Each core has a unique quantum identification and is keyed to a specific AI matrix. Once the signatures are keyed during the initialization phase the node's ECC is forced into an entangled state with the ECC within the central core. After initialization the nodes effectively operate as a single core over any distance via the subspace particle entanglement nearly instantaneously. This mutually entangled network with the central core is called the node array.
Intra-Array Transit
An AI matrix need not physically transit between component nodes within the node array as all nodes are synchronous. Thus the node array operates at all locations simultaneously.
Inter-Array Transit
It is not possible to transfer an AI matrix into a node array initialized for another AI. Two AI's would communicate via conventional means or via existing communications channels. A multinodal core must be initialized to one, and only one, AI matrix. Two AI's cannot exist within the same node array.
Extra-Array Transit
An AI may use communications systems to transit out of their initialized node array. The system is designed to keep the matrix synchronous within the array and allow operation of a single AI within disconnected nodes. In some situations it may be required for an AI to transfer their program outside of an array, though any core can do this the ATP attempts to establish a link from the full core to the transit destination if at all possible. The introduction of a system not within the node array will cause lag issues sufficient to desynchronize the remote AI operating outside the node array from the array.
Desynchronization
When the AI matrix in different physical locations goes out of sync with copies at other locations the ATP creates multiple virtual node instances within the central core. These virtual nodes exist as gateways for the AI matrix to relay information to and from the node array, primarily from systems outside the array. By creating virtual nodes the array can have two instances of the same AI matrix running within a singular core while one stays within the array the other is connected outside the array to an external system. This buffer between the array and non-array instances allow the AI matrix remain stable even with different parts are running at different speeds. Any core attached directly, via physical connection, to another computer system should not suffer desynchronization unless the connection quality or stability is low.
Split Array
As a special case of desynchronization an array can be split. If the various cores are forced out of the range of entanglement. This happens during extra-dimensional or extra-temporal travel. During these kinds of transit the system is unable to maintain an entangled synchronous link between parts in different dimensional or temporal zones. Each split of the array functions as it own array. Entanglement persists even when parts of the array are out of range, though the node network is unable to operate synchronously until the formerly split parts are brought back in range.
Temporal Synchronicity
A split array has a danger that when two parts of the array come within entangled range but are outside of their origin time zone they can resync, so long as the node array components are functioning. This happens when a node array is split due to extra-temporal travel and one part of the array is at a different point along the same timeline as another part which both exist at the same time and within the same region of space-time. While normally the array will not resynchronize with an array from a previous point along its own timestream to avoid causality problems, this is a safety protocol in place of the ATP and is not an given. If this safety protocol is disabled in some way or fails to successfully maintain a split array, both arrays will rejoin into a complete array.
The protocol is designed to handle extra-temporal information exchange but problems can develop. Both sides of the split array do restore their singular array state, leaving the network segmented in software. Should this protection fail or be disabled a causality problem could develop. The previous stage of the same AI could become aware of future events which then could change the outcome of those same future events.