Joint resource allocation method for underwater cloud-edge collaborative network

Abstract

Disclosed in the present invention is a joint resource allocation method for an underwater cloud-edge collaborative network. The method comprises the following steps: according to geographic locations, dividing a water region into several rectangular grids having equal areas; obtaining a Hamiltonian loop according to an APPROX-TSP-TOUR algorithm; calculating a system delay and an energy loss, wherein the system delay comprises: a service transmission delay, a service calculation delay and a service movement delay, and the energy loss comprises an underwater moving-ahead energy loss, calculation loss and communication transmission loss of an underwater mobile submersible vehicle; and constructing a resource allocation model in which both the system delay and the energy loss are taken into consideration, and performing calculation by means of linear programming, so as to obtain an optimal resource allocation strategy. In the present invention, a regional collection strategy is specified for underwater sensor data, and a collection path is programmed, such that the efficiency is improved, the energy is saved, and the phenomenon of uneven resource allocation is avoided; and a resource allocation model in which both a system delay and an energy loss are taken into consideration is constructed, such that a relatively short delay is ensured while the energy loss is reduced.

Description

[0002] The present disclosure relates to the field of edge computing, and in particular, to a joint resource allocation method for an underwater cloud-edge collaborative network. BACKGROUND

[0003] With the continuous innovation of technologies such as the Internet of Things (loT) and big data, the era of the Internet of Everything (loE) has arrived. Such a sustained and rapid growth in a data volume has driven evolution of an entire computing model, but imposes a higher requirement for data storage and processing technologies, especially for real-time management and intelligent analysis of businesses at a network edge. However, due to inherent drawbacks of traditional cloud computing, such as insufficient real-time performance, a limited bandwidth, a high energy consumption, a lack of mobility support and location awareness, its key technologies can no longer efficiently process massive data generated by an edge device. Therefore, edge computing has emerged as a new computing model. By extending a computing capability from a cloud data center to the network edge, the edge computing makes local data analysis and management more flexible and controllable, which is fundamental for implementing the loT.

[0004] Nowadays, an loT device is widely introduced into a marine environment, primarily for marine monitoring and exploration. These efforts mainly focus on a perception layer and have played an unparalleled role in promoting development of an underwater sensor network in the ocean. As a quantity of sensor nodes increases, a linearly growing centralized cloud computing capability can no longer meet an explosive growth of massive marine data (including marine measured data, marine remotely-sensed data, marine model data, and the like). A burden of transmitting massive data from an underlying device in the underwater sensor network to a cloud center is increasingly heavy. Therefore, the emergence of the edge computing has played a buffering role for this issue by processing data at the network edge. This can reduce response time for an underlying request, enhance a battery life of the underlying device, reduce a network bandwidth, and ensure data security and privacy. Thus, introducing the edge computing is a significant measure for marine development and national sustainable development.

[0005] As an emerging computing paradigm, the edge computing pushes computing, storage, network, and other elastic resources to the network edge, providing intelligent computing services nearby and addressing problems like a high delay and poor mobility in the cloud computing. Compared with the cloud data center, a limited cloud resource of an edge is a key bottleneck for improving performance of the edge computing, especially under uneven spatiotemporal distribution of computing loads and resources. Therefore, how to achieve rational scheduling of an end-edge-cloud computing resource and rapid and efficient execution of a computing task by designing an efficient resource optimization and configuration strategy is currently a major research focus and challenge in the field of the edge computing.

[0006] Unlike the rapid development of the edge computing in a land-based scenario dominated by wireless communication, research on applying the edge computing to a special application scenario, namely an underwater environment, is still in its infancy. The underwater environment is complex and changeable, with difficulties in energy replenishment, a limited communication distance, and a large delay. There are very few edge computing offloading strategies and resource optimization algorithms tailored to this specific application scenario. Directly applying an algorithm designed for the land-based scenario to the underwater environment fails to conduct a specific analysis for a specific problem and is unlikely to yield an ideal result. Meanwhile, most existing strategies for collecting underwater sensor data only focus on planning a collection path and rarely incorporate both an energy loss and a system delay into an optimization objective. SUMMARY

[0007] With respect to the deficiencies of the prior art, the present disclosure provides a joint resource allocation method for an underwater cloud-edge collaborative network.

[0008] Specific technical solutions are as follows:

[0009] A joint resource allocation method for an underwater cloud-edge collaborative network includes following steps:

[0010] step 1: based on a geographical location, dividing a water region into multiple rectangular grids with an equal area, where each of the multiple rectangular grids is provided with a same quantity of underwater sensors, and has an underwater mobile submersible vehicle for data collection independently;

[0011] step 2: obtaining, according to an Approximate Traveling Salesman Problem Tour (APPROX-TSP-TOUR) algorithm, a Hamiltonian loop constituted by a shore-based server and the underwater sensor in any one of the multiple rectangular grids, namely a movement route of a single underwater mobile submersible vehicle;

[0012] step 3: calculating a system delay and an energy loss, where the system delay includes: a service transmission delay, a service calculation delay, and a service movement delay; the service transmission delay is a sum of a transmission delay caused by underwater acoustic communication between the underwater sensor and the underwater mobile submersible vehicle, and a transmission delay caused by wireless communication between the underwater mobile submersible vehicle and the shore-based server; the service calculation delay is a calculation delay of data offloaded by the underwater mobile submersible vehicle to the shore-based server at the shore-based server; the service movement delay is time taken by the underwater mobile submersible vehicle to traverse the Hamiltonian loop; and the energy loss includes: an underwater moving-ahead energy loss, a calculation loss, and a communication transmission loss of the underwater mobile submersible vehicle; and

[0013] step 4: constructing, based on the system delay and the energy loss, a resource allocation model in which both the system delay and the energy loss are taken into consideration, performing linear programming to solve the resource allocation model in which both the system delay and the energy loss are taken into consideration, and obtaining an optimal resource allocation strategy.

[0014] Further, in the step 3, an expression of the service transmission delay is as follows:

[0015] where Dtot represents a total amount of data generated by the underwater sensor in the rectangular grid, Cacou represents a capacity of an underwater acoustic channel, Wacou represents a width of an underwater acoustic frequency band, represents an underwater ^acou acoustic signal-to-noise ratio, Dedge represents an amount of data processed by the underwater mobile submersible vehicle, Cwire represents a capacity of a wireless channel, Wwire 5 . represents a width of a wireless frequency band, and represents a wireless signal-to-noise Nwire ratio;

[0016] an expression of the service calculation delay is as follows: = 7, , / csssss

[0017] where fcioud represents a central processing unit (CPU) cycle required by the shore-based server to process 1-bit data each time; and

[0018] an expression of the service movement delay is as follows: £ 7* 3ZZ —* mow *

[0019] where L represents a path length of the Hamiltonian loop, and V represents a moving speed of the underwater mobile submersible vehicle.

[0020] Further, in the step 3, an expression of the underwater moving-ahead energy loss of the underwater mobile submersible vehicle is as follows:

[0021] where Emove represents an energy loss each time the underwater mobile submersible vehicle moves one meter;

[0022] an expression of the calculation loss of the underwater mobile submersible vehicle is as follows:

[0023] where Ecomp represents an energy loss of processing the 1-bit data each time; and

[0024] the communication transmission loss of the underwater mobile vehicle submersible vehicle includes an energy loss Eer^e edae for reception in the underwater acoustic communication between the underwater sensor and the underwater mobile submersible vehicle and an energy loss E?^ cloud for transmission in the wireless communication between the underwater mobile submersible vehicle and the shore-based server, that are expressed as follows: = X ; sens x

[0025] where Erece represents an energy loss of receiving the 1-bit data each time, and Esend represents an energy loss of sending the 1-bit data each time.

[0026] Further, in the step 4, an expression of the resource allocation model in which both the system delay and the energy loss are taken into consideration is as follows: arpminF T + rend^ed9e + redUe^ + Tdoud , , F?dge aiginm^ 1 move 1 1 tran ' 1 tran r ' comp ' 'W^move T ‘-‘comp Dedge cend—edge „edge—cloudy-. + ^rece + ^send 'J + bs8w + ■“sen^ 1' . mow ’ A >0;

[0027] where Etot represents total energy of the underwater mobile submersible vehicle.

[0028] Further, in the step 2, a specific process of solving the APPROX-TSP-TOUR algorithm is as follows:

[0029] step 2.1: using the underwater sensor and the shore-based server as nodes, and connecting the nodes to each other to form a complete undirected graph;

[0030] step 2.2: using a Euclidean distance as a cost function between two of the nodes;

[0031] step 2.3: obtaining a minimum spanning tree (MST) T through calculation by using an MST-PRIM algorithm;

[0032] step 2.4: performing preorder traversal on the generated MST T, and recording only a sequence of nodes visited for the first time; and

[0033] step 2.5: connecting the underwater sensor and the shore-based server sequentially based on the sequence of nodes obtained in the step 2.4, and obtaining the Hamiltonian loop.

[0034] The present disclosure has following beneficial effects:

[0035] (1) Based on a divide-and-conquer concept, the present disclosure determines a regional collection strategy for underwater sensor data, thereby avoiding uneven resource allocation.

[0036] (2) The present disclosure plans a movement path of an underwater mobile submersible vehicle, and obtains a Hamiltonian loop, which improves efficiency and conserves energy.

[0037] (3) The present disclosure constructs a resource allocation model in which both a system delay and an energy loss are taken into consideration, ensuring a low delay while reducing the energy loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] FIG. lisa flowchart of a method according to the present disclosure; and

[0039] FIG. 2 is flowchart of an APPROX-TSP-TOUR algorithm. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] The present disclosure is described in detail with reference to accompanying drawings and preferred embodiments, to make the objective and effect of the present disclosure clearer. The present disclosure is described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present disclosure and are not intended to limit the present disclosure.

[0041] As shown in FIG. 1, a joint resource allocation method for an underwater cloud-edge collaborative network includes following steps:

[0042] Step 1: Respectively plan an underwater sensor network, an underwater mobile submersible vehicle formation, and a shore-based server as an end, an edge, and a cloud in an edge computing model, and based on a geographical location, divide a water region into multiple rectangular grids with an equal area, where each of the multiple rectangular grids is provided with a same quantity of underwater sensors, and has an underwater mobile submersible vehicle for data collection independently.

[0043] Step 2: Obtain, according to an APPROX-TSP-TOUR algorithm, a Hamiltonian loop constituted by the shore-based server and the underwater sensor in any one of the multiple rectangular grids, namely a movement route of a single underwater mobile submersible vehicle.

[0044] As shown in FIG. 2, a specific process of solving the APPROX-TSP-TOUR algorithm is as follows:

[0045] Step 2.1: Use the underwater sensor and the shore-based server as nodes, and connect the nodes to each other to form a complete undirected graph.

[0046] Step 2.2: Use a Euclidean distance as a cost function between two of the nodes.

[0047] Step 2.3: Obtain an MST T through calculation by using an MST-PRIM algorithm.

[0048] Step 2.4: Perform preorder traversal on the generated MST T, and record only a sequence of nodes visited for the first time.

[0049] Step 2.5: Connect the underwater sensor and the shore-based server sequentially based on the sequence of nodes obtained in the step 2.4, and obtain the Hamiltonian loop.

[0050] Step 3 : Calculate a system delay and an energy loss.

[0051] A service transmission delay is obtained by adding up a transmission delay Te^dn edge caused by underwater acoustic communication between the underwater sensor and the underwater mobile submersible vehicle, and a transmission delay Tetdg^e cloud caused by wireless communication between the underwater mobile submersible vehicle and the shore-based server are added up, and is expressed as follows:

[0052] where Dtot represents a total amount of data generated by the underwater sensor in the rectangular grid, Cacou represents a capacity of an underwater acoustic channel, represents a width of an underwater acoustic frequency band, -2^ represents an underwater ^acou acoustic signal-to-noise ratio, Dedge represents an amount of data processed by the underwater mobile submersible vehicle, Cwire represents a capacity of a wireless channel, Wwire represents a width of a wireless frequency band, and represents a wireless signal-to-noise ratio.

[0053] A service calculation delay T^p is obtained based on a calculation delay of data offloaded from the underwater mobile submersible vehicle to the shore-based server at the shore-based server, and is expressed as follows: * “•

[0054] where fciouct represents a CPU cycle required by the shore-based server to process 1 -bit data each time.

[0055] A service movement delay Tmove is obtained based on time taken by the underwater mobile submersible vehicle to traverse the Hamiltonian loop, and is expressed as follows: Tmove

[0056] where L represents a path length of the Hamiltonian loop, and V represents a moving speed of the underwater mobile submersible vehicle.

[0057] The system delay is obtained by adding up the service transmission delay, the service calculation delay, and the service movement delay. It should be noted that in a transmission process in which the underwater acoustic communication is performed between the underwater sensor and the underwater mobile submersible vehicle and the wireless communication is performed between the underwater mobile submersible vehicle and the shore-based server, a sending delay of a sending end and a processing delay of a receiving end are not taken into consideration.

[0058] Meanwhile, an underwater moving-ahead energy loss, a calculation loss, and a communication transmission loss of the underwater mobile submersible vehicle are added up to obtain the energy loss. It should be noted that only an energy loss generated by operation of the underwater mobile submersible vehicle in a system is taken into consideration, and energy losses of the shore-based server and the underwater sensor are not taken into consideration.

[0059] The underwater moving-ahead energy loss of the underwater mobile submersible vehicle is a moving-ahead energy loss Emove generated by a single underwater mobile submersible vehicle when traversing the Hamiltonian loop within its jurisdiction, and is expressed as follows: ' b -it.

[0060] where Emove represents an energy loss each time the underwater mobile submersible vehicle moves one meter.

[0061] The calculation loss of the underwater mobile submersible vehicle is an energy loss Ecomp °f processing data by the underwater mobile submersible vehicle, and is expressed as follows: s _ n v S'

[0062] where Ecomp represents an energy loss of processing the 1 -bit data each time.

[0063] The communication transmission loss of the underwater mobile vehicle submersible vehicle includes an energy loss E^e edge for reception in the underwater acoustic communication between the underwater sensor and the underwater mobile submersible vehicle and an energy loss E^a Cl°Ud for transmission in the wireless communication between the underwater mobile submersible vehicle and the shore-based server, which are expressed as follows:

[0064] where Erece represents an energy loss of receiving the 1 -bit data each time, and Esenci represents an energy loss of sending the 1 -bit data each time.

[0065] Step 4: Construct, based on the system delay and the energy loss and in combination with a weight parameter X, a resource allocation model in which both the system delay and the energy loss are taken into consideration, perform linear programming to solve the resource allocation model in which both the system delay and the energy loss are taken into consideration, and obtain an optimal resource allocation strategy. An expression of the resource allocation model in which both the system delay and the energy loss are taken into consideration is as follows: . ^end-edge , j,edge-cloud ' 1 tran 1 tran , Tcloud i yep ! ped9e , pend^edge ' 1 comp ' ^move ' ^comp ' ^rece argmin [T3 Dedge । r^edge cloudy. + b send J J A >0.

[0066] where Etot represents total energy of the underwater mobile submersible vehicle.

[0067] In the resource allocation model in which both the system delay and the energy loss are taken into consideration, any submersible vehicle in the underwater mobile submersible vehicle formation can determine its own Dedge based on a distribution of a sensor within its jurisdiction and the shore-based server, and must traverse all sensors within its jurisdiction.

[0068] The linear programming is performed to solve the resource allocation model in which both the system delay and the energy loss are taken into consideration. Energy and delay expressions are substituted into the expression of the resource allocation model in which both the system delay and the energy loss are taken into consideration, and it can be obtained that: “E?"K x D‘d” <+AL x + x x ^wire / Erece + ^Etot * Esenci + / swire\ Wwirelog? I+tW X ^wire / _________Dtot__________ । ^tot ^acoul°92 (1+ fcloud) 1 x ™acou' 0 <A.

[0069] It can be obtained that:

[0070] Values of the Emove, the re*® , the , and the Eec^p are calculated, that is, the optimal resource allocation strategy is obtained.

[0071] A person of ordinary skill in the art may understand that the above descriptions are only preferred embodiments of the present disclosure and are not intended to limit the present disclosure. Although the present disclosure is described in detail with reference to the foregoing embodiments, a person skilled in the art can still make modifications to the technical solutions described in the foregoing embodiments, or make equivalent replacement to some technical characteristics. Any modifications and equivalent substitutions made within the spirit and scope of the present disclosure should be included within the protection scope of the present disclosure.

Claims

1. A joint resource allocation method for an underwater cloud-edge collaborative network, comprising following steps:step 1: based on a geographical location, dividing a water region into multiple rectangular grids with an equal area, wherein each of the multiple rectangular grids is provided with a same quantity of underwater sensors, and independently has an underwater mobile submersible vehicle for data collection;step 2: obtaining, according to an Approximate Traveling Salesman Problem Tour (APPROX-TSP-TOUR) algorithm, a Hamiltonian loop constituted by a shore-based server and the underwater sensor in any one of the multiple rectangular grids, namely a movement route of a single underwater mobile submersible vehicle;step 3: calculating a system delay and an energy loss, wherein the system delay comprises: a service transmission delay, a service calculation delay, and a service movement delay; the service transmission delay is a sum of a transmission delay caused by underwater acoustic communication between the underwater sensor and the underwater mobile submersible vehicle, and a transmission delay caused by wireless communication between the underwater mobile submersible vehicle and the shore-based server; the service calculation delay is a calculation delay of data offloaded by the underwater mobile submersible vehicle to the shore-based server at the shore-based server; the service movement delay is time taken by the underwater mobile submersible vehicle to traverse the Hamiltonian loop; and the energy loss comprises: an underwater moving-ahead energy loss, a calculation loss, and a communication transmission loss of the underwater mobile submersible vehicle; andstep 4: constructing, based on the system delay and the energy loss, a resource allocation model in which both the system delay and the energy loss are taken into consideration, performing linear programming to solve the resource allocation model in which both the system delay and the energy loss are taken into consideration, and obtaining an optimal resource allocation strategy.

2. The joint resource allocation method for an underwater cloud-edge collaborative network according to claim 1, wherein in the step 3, an expression of the service transmission delay is as follows:wherein Dtot represents a total amount of data generated by the underwater sensor in the rectangular grid, Cacou represents a capacity of an underwater acoustic channel, Wacou represents a width of an underwater acoustic frequency band, represents an underwater acoustic signal-to-noise ratio, Dedge represents an amount of data processed by the underwater mobile submersible vehicle, Cwire represents a capacity of a wireless channel, Wwire $ .represents a width of a wireless frequency band, and represents a wireless signal-to-noise ^wireratio;an expression of the service calculation delay is as follows:'pcloud _i comp ~f iJ cloudwherein Ycloud represents a central processing unit (CPU) cycle required by the shore-based server to process 1 -bit data each time; andan expression of the service movement delay is as follows:T = --wherein L represents a path length of the Hamiltonian loop, and V represents a moving speed of the underwater mobile submersible vehicle.

3. The joint resource allocation method for an underwater cloud-edge collaborative network according to claim 2, wherein in the step 3, an expression of the underwater moving-ahead energy loss of the underwater mobile submersible vehicle is as follows:wherein Emove represents an energy loss each time the underwater mobile submersible vehicle moves one meter;an expression of the calculation loss of the underwater mobile submersible vehicle is as follows:-MOT??wherein Ecomp represents an energy loss of processing the 1-bit data each time; andthe communication transmission loss of the underwater mobile vehicle submersible vehiclecomprises an energy loss for reception in the underwater acoustic communicationbetween the underwater sensor and the underwater mobile submersible vehicle and an energyloss for transmission in the wireless communication between the underwatermobile submersible vehicle and the shore-based server that are expressed as follows:.edge—ckadwherein Erece represents an energy loss of receiving the 1-bit data each time, and Esenct represents an energy loss of sending the 1-bit data each time.

4. The joint resource allocation method for an underwater cloud-edge collaborative networkaccording to claim 3, wherein in the step 4, an expression of the resource allocation model inwhich both the system delay and the energy loss are taken into consideration is as follows:aramint T i 'pend edge . ^edge cloud . rpcloud dlgllllll[ j move 1 tran ' 1 tran ' 1 compDedge. . r^edge , ^,end—edgemove comp ^rece। r^edge cloudy + ^send J-I0 — Dedge <Dtot-, A >0;wherein Etot represents total energy of the underwater mobile submersible vehicle.

5. The joint resource allocation method for an underwater cloud-edge collaborative network according to claim 1, wherein in the step 2, a specific process of solving the APPROX-TSP-TOUR algorithm is as follows:step 2.1: using the underwater sensor and the shore-based server as nodes, and connecting the nodes to each other to form a complete undirected graph;step 2.2: using a Euclidean distance as a cost function between two of the nodes;step 2.3: obtaining a minimum spanning tree (MST) T through calculation by using an MST-PRIM algorithm;step 2.4: performing preorder traversal on the generated MST T, and recording only a sequence of nodes visited for the first time; andstep 2.5: connecting the underwater sensor and the shore-based server sequentially based on the sequence of nodes obtained in the step 2.4, and obtaining the Hamiltonian loop.INTERNATIONAL SEARCH REPORT International application No. PCT / CN2024 / 094378A. CLASSIFICATION OF SUBJECT MATTER H04W72 / 53(2023.01)i According to International Patent Classification (IPC) or to both national classification and IPC B. FIELDS SEARCHED Minimum documentation searched (classification system followed by classification symbols) IPC:H04W Documentation searched other than minimum documentation to the extent that such documents are included in the fields searched Electronic data base consulted during the international search (name of data base and, where practicable, search terms used) CNABS, CNTXT, CNKI, VEN, WOTXT, EPTXT, USTXT, CJFD: 7KT, St®, tEH WK water, sea, time delay, energy, loss, cloud collaboration, network C. DOCUMENTS CONSIDERED TO BE RELEVANT Category* Citation of document, with indication, where appropriate, of the relevant passages Relevant to claim No. PX CN 116234045 A (ZHEJIANG UNIVERSITY) 06 June 2023 (2023-06-06) claims 1-5 1-5 A A A CN 113891275 A (DALIAN UNIVERSITY OF TECHNOLOGY) 04 January 2022 (2022-01-04) entire document CN 112787920 A (XIAMEN UNIVERSITY) 11 May 2021 (2021-05-11) entire document WO 2021217942 Al (INSTITUTE OF OCEANOGRAPHIC INSTRUMENTATION, SHANDONG ACADEMY OF SCIENCES) 04 November 2021 (2021-11-04) entire document 1-5 1-5 1-5 A (SHI, Wenyu; ZHANG, Rui). MX (Non-official translation: Security-Aware Workflow Task Scheduling Strategy in Edge-Cloud Collaborative Computing)" KKKKKKKX (Journal of Changchun Normal University), No. 08, 20 August 2020 (2020-08-20), entire document 1-5 | | Further documents are listed in the continuation of Box C. | J | See patent family annex. * Special categories of cited documents: “T” later document published after the international filing date or priority “A” document defining the general state of the art which is not considered date and not in conflict with the application but cited to understand the to be of particular- relevance principle or theory underlying the invention “D” document cited by the applicant in the international application -‘X” document of particular relevance; the claimed invention cannot be “E" earlier application or patent but published on or after the international considered novel or cannot be considered to involve an inventive step filing date when the document is taken alone “L" document which may throw doubts on priority claim(s) or which is “Y” document of particular relevance; the claimed invention cannot be cited to establish the publication date of another citation or other considered to involve an inventive step when the document is special reason (as specified) combined with one or more other such documents, such combination “O” document referring to an oral disclosure, use, exhibition or other being obvious to a person skilled in the ait means document member of the same patent family “P” document published prior to the international filing date but later than the priority date claimed Date of the actual completion of the international search 15 July 2024 Date of mailing of the international search report 13 September 2024 Name and mailing address of the ISA / CN China National Intellectual Property Administration (ISA / CN) China No. 6, Xitucheng Road, Jimenqiao, Haidian District, Beijing 100088 Authorized officer Telephone No.INTERNATIONAL SEARCH REPORT Information on patent family membersInternational application No.PCT / CN2024 / 094378Patent document cited in search report Publication date (day / month / year) Patent family member) s) Publication date (day / month / year) CN 116234045 A 06 June 2023 None CN 113891275 A 04 January 2022 None CN 112787920 A 11 May 2021 None WO 2021217942 Al 04 November 2021 None

Images