Methods, terminals and network devices for computing task allocation and updating
The disclosure discloses a computing task allocation method, an updating method for computing task allocation, a terminal and a network device. When a computing task of a terminal is generated, computing task allocation is performed using at least one of a centralized mode, a distributed mode, or a hybrid mode; the computing task allocation includes communication resource allocation, computing resource allocation, and a task offloading decision; the above computing task allocation method subjected to dynamically updating according to a terminal state, a network state or a task state. Therefore, the compromise problem between overall system performance optimization and device fairness in a cloud-edge collaborative IoT system is solved.
1. A terminal, wherein the terminal is configured to implement the following operations:
generating a computing task for execution;
when computing task allocation is performed, by a network device, using a centralized mode, reporting features of the computing task and terminal energy consumption to the network device, and receiving an allocation result of the computing task allocation issued by the network device; wherein the features of the computing task comprising task load, amount of computation required by the computing task, task type, maximum tolerance latency of the task, and priority of the computing task;
when computing task allocation is performed using a distributed mode, performing the computing task allocation locally, to the terminal, according to the features of the computing task, unit price of computing resources, terminal energy consumption, and expected transmission latency; and uploading an allocation result of the computing task allocation to the network device in whole or in part to assist the network device in making a task allocation decision for one or more other computing tasks;
when computing task allocation is performed using a hybrid mode, performing the computing task allocation locally, to the terminal, according to an instruction, issued to a plurality of terminals including the terminal, from the network device, and uploading an allocation result, that is specific to the terminal, of the computing task allocation to the network device in whole or in part to assist the network device in performing the instruction;
the computing task allocation comprises a communication resource allocation, a computing resource allocation, and a task offloading decision;
determining whether to locally, to the terminal, compute or to offload the computing task according to task offloading decision; and when the computing task is determined to be offloaded, determining a transmission path of the task and transmitting the task to a target computing server on the allocated communication resource;
receiving a computation result for the computing task;
transmitting a task allocation result to the network device in whole or in part, helping the network device to grasp the computing resource distribution state of an entire system; and
the instruction of the network device specifically comprising a task offloading decision comprising determining a maximum number of tasks that can be offloaded or determining a maximum amount of computation of tasks that can be offloaded.
2. The terminal according to claim 1 , wherein the computing task allocation performed by the network device specifically comprises the following operations: performing communication resource allocation and computing resource allocation, and making a task offloading decision by the network device according to features of computing tasks, terminal energy consumption, terminal computing power reported by respective terminals, computing power of the network device, and network topology.
3. The terminal according to claim 1 , wherein the computing task allocation locally performed by the terminal specifically comprises the following operations: performing communication resource allocation, and making a task offloading decision by the terminal according to features of the computing tasks, unit price of computing resources, terminal energy consumption, terminal storage capacity and expected transmission latency.
4. The terminal according to claim 1 , wherein the task offloading decision is made taking into account one or more of the following factors:
features of a computing task, comprising: task load, amount of computation required by the task, task type, maximum tolerance latency of the task, and priority of the task;
computing power, comprising computing-capable devices in the network;
unit price of computing resources;
terminal storage capacity;
terminal energy consumption;
expected transmission latency, comprising: transmission latency for a task to be uploaded to the network device, transmission latency for a computation result to be returned to the terminal, and relay latency; and
network topology.
5. The terminal according to claim 1 , wherein the communication resource allocation comprises communication resources required to transmit the computing task to the target computing server.
6. The terminal according to claim 5 , wherein the communication resource allocation further comprises communication resources required for receiving a computation result.
7. The terminal according to claim 1 , wherein the terminal is further configured to implement the following operations: performing dynamic updating on the computing task allocation according to a terminal state, a network state or a task state in one of a distributed mode or a hybrid mode, or between the distributed mode or the hybrid mode.
8. The terminal according to claim 7 , wherein the terminal state comprises mobility, location and an energy consumption state; the network state comprises a change in network topology, a network congestion situation, release and occupancy situations of computing resources in a network; and the task state comprises a task queue state, a change trend of task types.
9. The terminal according to claim 7 , wherein the dynamic updating is periodic updating or aperiodic updating, and the aperiodic updating comprises event-based updating or triggering-based updating.
10. A terminal, wherein the terminal is configured to implement the following operations:
generating a computing task for execution;
when computing task allocation is performed, by a network device, using a centralized mode, reporting features of the computing task and terminal energy consumption to the network device, and receiving an allocation result of the computing task allocation issued by the network device; wherein the features of the computing task comprising task load, amount of computation required by the computing task, task type, maximum tolerance latency of the task, and priority of the computing task;
when computing task allocation is performed using a distributed mode, performing the computing task allocation locally, to the terminal, according to the features of the computing task, unit price of computing resources, terminal energy consumption, and expected transmission latency; and uploading an allocation result of the computing task allocation to the network device in whole or in part to assist the network device in making a task allocation decision for one or more other computing tasks;
when computing task allocation is performed using a hybrid mode, performing the computing task allocation locally, to the terminal, according to an instruction, issued to a plurality of terminals including the terminal, from the network device, and uploading an allocation result, that is specific to the terminal, of the computing task allocation to the network device in whole or in part to assist the network device in performing the instruction;
the computing task allocation comprises a communication resource allocation, a computing resource allocation, and a task offloading decision;
determining whether to locally, to the terminal, compute or to offload the computing task according to different solutions for the task offloading decision; and when the computing task is determined to be offloaded, determining a transmission path of the task and transmitting the task to a target computing server using the allocated communication resource;
receiving a computation result for the computing task;
transmitting a task allocation result to the network device in whole or in part, helping the network device to grasp the computing resource distribution state of an entire system; and
the instruction of the network device specifically comprising a hybrid mode specific solution for the task offloading decision.
11. The terminal according to claim 10 , wherein the computing task allocation performed by the network device specifically comprises the following operations: performing communication resource allocation and computing resource allocation, and making a task offloading decision by the network device according to features of computing tasks, terminal energy consumption, terminal computing power reported by respective terminals, computing power of the network device, and network topology.
12. The terminal according to claim 10 , wherein the computing task allocation locally performed by the terminal specifically comprises the following operations: performing communication resource allocation, and making a task offloading decision by the terminal according to features of the computing tasks, unit price of computing resources, terminal energy consumption, terminal storage capacity and expected transmission latency.
13. The terminal according to claim 10 , wherein the task offloading decision is made taking into account one or more of the following factors:
features of a computing task, comprising: task load, amount of computation required by the task, task type, maximum tolerance latency of the task, and priority of the task;
computing power, comprising computing-capable devices in the network;
unit price of computing resources;
terminal storage capacity;
terminal energy consumption;
expected transmission latency, comprising: transmission latency for a task to be uploaded to the network device, transmission latency for a computation result to be returned to the terminal, and relay latency; and
network topology.
14. The terminal according to claim 10 , wherein the communication resource allocation comprises communication resources required to transmit the computing task to the target computing server.
15. The terminal according to claim 14 , wherein the communication resource allocation further comprises communication resources required for receiving a computation result.
16. The terminal according to claim 10 , wherein the terminal is further configured to implement the following operations: performing dynamic updating on the computing task allocation according to a terminal state, a network state or a task state in one of a distributed mode or a hybrid mode, or between the distributed mode or the hybrid mode.
17. The terminal according to claim 16 , wherein the terminal state comprises mobility, location and an energy consumption state; the network state comprises a change in network topology, a network congestion situation, release and occupancy situations of computing resources in a network; and the task state comprises a task queue state, a change trend of task types.
18. The terminal according to claim 16 , wherein the dynamic updating is periodic updating or aperiodic updating, and the aperiodic updating comprises event-based updating or triggering-based updating.