Network virtualization and resource allocation for the internet of things
A system and method for network virtualization and resource allocation, including storing one or more received network requests in a request table, and updating at least one of a flow table, a waiting list table, or a candidate group map based on the one or more received network requests. The updating includes monitoring a transmission progress of each of one or more flows in a network of interconnected computing devices and moving completed flows from the flow table to a success list, moving requests in the waiting list table which have reached an attempt threshold from the flow table to a fail list, and compiling any residual requests in the waiting list with new requests to generate a new request table. A deterministic request allocation and/or an optimizing request allocation is performed based on the new request table.
1. A computer-implemented method for network virtualization and resource allocation, comprising:
storing one or more received network requests in a request table;
managing, using a hardware processor coupled to a memory, network flows and requests by dynamically updating a flow table, a waiting list table, and a candidate group map based on the one or more received network requests, the managing further comprising:
monitoring a transmission progress of each of one or more flows in a network of interconnected computing devices, and moving completed flows from the flow table to a success list;
moving requests in the waiting list table which have reached an attempt threshold from the flow table to a fail list;
compiling any residual requests in the waiting list with new requests to generate a new request table; and
minimizing bandwidth usage by enabling the network to serve a plurality of requests simultaneously by sharing a single copy of bandwidth resources in central nodes using multicasting; and
fulfilling at least one of a deterministic request allocation or an optimizing request allocation based on the new request table,
wherein if no central node is determined to be available to accommodate a highly ranked request, lower ranked in-progress flows are iteratively suspended until the highly ranked request is accommodated.
2. The method as recited in claim 1 , wherein the deterministic request allocation further comprises sorting the requests based on at least one of request priorities or values of profit ratios.
3. The method as recited in claim 1 , wherein the multicasting further comprising serving a request without consuming any additional bandwidth in one or more central nodes by locating one or more reusable flows carrying a requested service at a same transmission progress.
4. The method as recited in claim 1 , further comprising suspending one or more in-progress flows to free-up bandwidth resources if higher priority incoming requests are received.
5. The method as recited in claim 4 , further comprising generating a suspendable request list.
6. The method as recited in claim 5 , wherein if the suspendable list is not empty, a lowest priority in-progress flow from the suspendable list is suspended to enable space for the higher priority incoming requests.
7. A system for network virtualization and resource allocation, comprising:
a hardware processor coupled to a storage medium for storing one or more received network requests in a request table, the hardware processor being configured to:
manage network flows and requests by controlling adjustment, using an updater, of a flow table, a waiting list table, and a candidate group map based on the one or more received network requests, the updater being further configured to:
monitor a transmission progress of each of one or more flows in a network of interconnected computing devices and move completed flows from the flow table to a success list;
move requests in the waiting list table which have reached an attempt threshold from the flow table to a fail list;
compile any residual requests in the waiting list with new requests to generate a new request table; and
minimize bandwidth usage by enabling the network to serve a plurality of requests simultaneously by sharing a single copy of bandwidth resources in central nodes using a multicaster; and
fulfill, using a request allocator, at least one of a deterministic request allocation or an optimizing request allocation based on the new request table,
wherein if no central node is determined to be available to accommodate a highly ranked request, lower ranked in-progress flows are iteratively suspended until the highly ranked request is accommodated.
8. The system as recited in claim 7 , wherein the deterministic request allocation further comprises sorting the requests based on at least one of request priorities or values of profit ratios.
9. The system as recited in claim 7 , wherein the multicaster serves a request without consuming any additional bandwidth in one or more central nodes by locating one or more reusable flows carrying a requested service at a same transmission progress.
10. The system as recited in claim 7 , further comprising a controller for suspending one or more in-progress flows to free-up bandwidth resources if higher priority incoming requests are received.
11. The system as recited in claim 10 , wherein a table generator is configured to generate a suspendable request list based on the one or more in-progress flows suspended by the controller.
12. The system as recited in claim 11 , wherein if the suspendable request list is not empty, a lowest priority in-progress flow from the suspendable request list is suspended to enable space for the higher priority incoming requests.