KR102560159B1 - Network slice allocation in collaborative teleoperation system - Google Patents

Abstract

An apparatus for allocating network slices in a cooperative remote control system is provided. The cooperative remote control system operates on a network providing a plurality of haptic connections through which respective cooperative remote controllers are controlled. The provided apparatus includes a haptic connectivity monitoring module that measures a Quality of Service (QoS) value corresponding to each of a plurality of haptic connections, detects whether each of the plurality of haptic connections is unusual or normal based on the measured QoS value, and if an anomalous one of the plurality of haptic connections is detected, determines a QoS constraint based on the measured QoS value of one or more normal ones of the plurality of haptic connections, and a slice of the network that satisfies the QoS constraint. and an admission control module that identifies a set of slices in the network, and a slice assignment module that selects particular slices within the identified set of slices of the network for assignment to an unusual haptic connection.

Description

Network slice allocation in cooperative remote control system {NETWORK SLICE ALLOCATION IN COLLABORATIVE TELEOPERATION SYSTEM}

The present disclosure relates to network slice allocation in a collaborative teleoperation system.

Tactile Internet is an emerging technology in the field of Human-to-Machine (H2M) interaction, and is used to meet the ultra-low-latency, high reliability, high scalability, availability and security requirements of tactile applications. This Internet network provides a new paradigm shift from a content delivery network to a skills-set delivery network. For example, a cooperative remote control system may operate in such a networked environment for tactile applications. The cooperative remote control system includes a master domain in which a plurality of operators or devices thereof exist, and a slave domain in which a plurality of teleoperators exist, and these operators provide haptic control to the slave domains, and the remote controllers aim to complete the task in synchronization so that they can jointly perform a common task in the same place.

However, with respect to the delay requirements of network environments such as the tactile Internet, generally tactile applications target a round-trip delay of 1 ms, which is one of the challenges. In research and development related to network coding and haptic communication protocols, upgrading traditional network infrastructure and technology is complex, and testing new proposals takes a lot of time and resources. Regarding resource allocation or provisioning solutions, ultra-low latency requirements are not reliably guaranteed, and only limited improvements have been made so far.

Allocation of network slices in a cooperative remote control system is disclosed in this document.

In an example, an apparatus is provided for network slice allocation in a collaborative teleoperation system, wherein the cooperative telecontrol system operates on a network providing a plurality of haptic connections through which respective cooperative remote controllers are controlled, the apparatus comprising: measuring a Quality of Service (QoS) value corresponding to each of the plurality of haptic connections, and determining whether each of the plurality of haptic connections is anomalous or normal based on the measured QoS value. a haptic connectivity monitoring module that detects and, if an anomalous one of the plurality of haptic connections is detected, determines a QoS constraint based on a measured QoS value of one or more normal ones of the plurality of haptic connections; an admission control module that identifies a set of slices in the network that satisfy QoS constraints; and a slice allocation module that selects particular slices within the identified set of slices of the network for assignment to an unusual haptic connection.

The foregoing summary is provided to introduce some aspects in a simplified form that are further described below in the Detailed Description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to delineate the scope of the claimed subject matter. Furthermore, the claimed subject matter is not limited to implementations that provide any or all of the advantages discussed herein.

The present disclosure provides a solution to address ultra-low latency requirements with synchronization of cooperative remote controllers in cooperative remote control applications.

According to the present disclosure, ultra-low latency requirements can be satisfied through the operation of dynamic network slicing, and similar Quality of Service (QoS) values can be guaranteed between haptic connections to which cooperative remote controllers are connected in a cooperative remote control application.

According to the present disclosure, dynamic allocation of network slices is enabled along with continuous monitoring of haptic connectivity.

1 is a block diagram illustrating an example of network slice allocation in an exemplary cooperative remote control system.
FIG. 2 shows an example of a network slice provided on the core side of the network of FIG. 1 .
3 is a block diagram showing an example of the network controller of FIG. 1;
FIG. 4 is a flow diagram showing an exemplary haptic connectivity monitoring process performed by the network controller of FIG. 1 .
FIG. 5 is a diagram for explaining an exemplary SLA monitoring for measuring a QoS value in the haptic connectivity monitoring process of FIG. 4 .
6 is a flow diagram showing an exemplary admission control process performed by the network controller of FIG. 1;
7 is a flow diagram showing an exemplary slice allocation process performed by the network controller of FIG. 1;

Various terms used in the present disclosure are selected from the terminology of common terms in consideration of functions in this document, which may be recognized differently according to the intention of those skilled in the art, conventions, or the emergence of new technologies. In certain instances, some terms may be given meanings as set forth in the Detailed Description. Therefore, the terms used in this document should be defined consistently with the meanings of the terms in the context of the present disclosure, not simply by their names.

In this document, the terms "comprise", "have", etc. are used when specifying the presence of elements listed below, such as certain features, numbers, steps, operations, components, information, or combinations thereof. Unless otherwise indicated, these terms and variations thereof are not intended to exclude the presence or addition of other elements.

As used herein, the terms "first", "second", etc. are intended to identify several resembling elements. Unless otherwise stated, such terms are not intended to impose limitations on these elements or on the specific order of their use, but are merely used to refer to several elements separately. For example, an element may be referred to by the term "first" in one example, while the same element may be referred to by a different ordinal number, such as "second" or "third" in another example. In such instances, these terms are not intended to limit the scope of this disclosure. Further, use of the term “and/or” in a list of multiple elements includes all possible combinations of any one or more of the listed items, including any one or more of these items. Further, expressions in the singular form include the meanings in the plural form unless the context clearly dictates otherwise.

Certain examples of the present disclosure will now be described in detail with reference to the accompanying drawings. However, this disclosure may be embodied in many different forms and should not be construed as being limited to the examples set out in this document. Rather, these examples are given to provide a better understanding of the scope of the present disclosure.

1 is a block diagram illustrating an example of network slice allocation in an exemplary cooperative remote control system 100.

In the example of FIG. 1 , the cooperative remote control system 100 includes a master domain 110, a slave domain 120, and a network domain 150 interposed therebetween.

In the illustrated example, master domain 110 includes a plurality of master devices 115-1, 115-2, ..., 115-n (which may be individually or collectively referred to below with reference numeral "115"). For example, each master device 115 may be a device used by a human operator (which may also be referred to below as a user), such as a haptic glove or other type of haptic device that may be worn by a human operator.

In the illustrated example, the slave domain 120 includes a plurality of slave devices (or remote controllers) 125-1, 125-2, ..., 125-n (which may be individually or collectively referred to below as reference numeral "125"). For example, each remote control 125 may be a robotic hand (which may be a robotic end-effector with a two, three, four, five, or other multi-finger mechanism, such as a multi-fingered robotic gripper), another robotic device, or another type of device. These slave devices 125-1, 125-2, ..., 125-n may be, for example, cooperating remote controllers controlled to perform a common task in one place.

In the illustrated example, network domain 150 includes network 151 on which cooperative remote control system 100 operates, which may be any suitable type of communication network, e.g., a network having an architecture that supports Internet Protocol (IP), such as a 5G network. For example, the network 151 may include a core network and may further include an access network that communicates with the core network. Network 151 may include various network entities for implementing various network functions and/or services. For example, as shown in FIG. 1 , network 151 may include a network controller 155 on the core network side. In a specific example, several network entities in network 151, such as network controller 155, may support a Software-Defined Networking (SDN) architecture.

In the illustrated example, network 151 provides a connection (also referred to herein as a “haptic connection”) between device 115 in master domain 110 and remote controller 125 in slave domain 120 so that these devices 115, 125 communicate with each other via a haptic connection established therebetween. For example, each of the plurality of remote controllers 125-1, 125-2, ..., 125-n in the slave domain 120 may be connected to a corresponding one of the plurality of devices 115-1, 115-2, ..., 115-n in the master domain 110 through different haptic connections.

In the illustrated example, remote control 125 is controlled via its haptic connection. For example, the master device 115 may provide a signal for haptic control, for example, a command signal generated by the master device 115 in response to a user's finger movement (which may include, for example, kinesthetic data) to the remote controller 125 through a haptic connection. Then, the remote controller 125 may perform an operation instructed by the haptic control signal. Remote controller 125 may also provide a haptic feedback signal (which may include, e.g., sensor data obtained by a sensor disposed on remote controller 125) to master device 115, e.g., as a response to the haptic control, via a haptic connection. The haptic feedback can then be processed by the master device 115 for tactile presentation, allowing the user to experience a sense of realism and immersion.

In some example implementations, network 151 can have a dynamically reconfigurable network architecture in which network slicing can be applied. In such example implementations, network 151 may be divided into multiple slices. Each slice of network 151 may be characterized by a combination of network functions that may be instantiated for services and/or applications supported by network 151 (eg, cooperative remote control applications). For example, each slice of network 151 may be associated with different performance metrics and/or service requirements. Accordingly, one of a plurality of slices of the network 151 may be dynamically allocated for data transfer between the master device 115 and the remote controller 125 . As such, a slice of network 151 is a logical network built on a shared physical infrastructure. Such network slicing may be enabled by Network Function Virtualization (NFV) technology in addition to SDN. NFV allows the use of common hardware for cost-effective implementation of different network functions, while SDN separates the control plane from the data plane to allow easier network management.

In some example implementations, information about slices configured for network 151 may be stored in storage included within network controller 155 or in an external repository accessible from network controller 155, e.g., in the form of a database.

In some example implementations, a slice of network 151 may include a collection of Virtualized Network Functions (VNFs). In a particular example, a slice of network 151 may be characterized by a number of ordered chains of VNFs, referred to as a Service Function Chain (SFC). As an example, SFC can be configured based on NFV and SDN. In a particular example, mapping such service chains for users may be performed on the core network side of network 151 . Additionally, at the access side of network 151, network slicing may be performed, such as to select a base station (BS) for a user.

2 shows an example of a network slice 200 provided on the core side of the network 151 . In the illustrated example, network slice 200 includes an SFC of three VNFs arranged in a specific order, denoted “Slice1: VNF1-VNF2-VNF3”. For example, the VNFs in the chain can each be a virtual computation function, a virtual storage function, and a virtual networking function. In addition, a specific combination of VNFs resides in the network slice 200 and an executable virtual machine (VM) may be provided. For example, FIG. 2 shows that network slice 200 includes the following VMs: VM1 with VNF1 deployed, VM2 with VNF2 deployed, VM3 with VNF1 and VNF2 deployed, and VM4 with VNF2 and VNF3 deployed.

3 is a block diagram showing an example of network controller 155 .

In the example of FIG. 3 , the network controller 155 includes an admission control module 310, a slice allocation module 320 and a haptic connectivity monitoring module 330. For example, network controller 155 may perform configuration, instantiation, management, and other control of slices of network 151 . In a particular example, network controller 155 may be an SDN-based controller deployed on the core side of network 151 . Other example implementations of network controller 155 are also contemplated. For example, network controller 155 may also include additional components not shown and/or may include some but not all of the components listed with respect to FIG. 1 .

In the illustrated example, admission control module 310 identifies a set of network slices corresponding to a given request, and slice assignment module 320 selects a particular network slice for assignment from among the identified set of network slices. For example, admission control module 310 may receive a user request for a particular level of service from master device 115 (e.g., provided via a haptic connection between master device 115 and remote controller 125) and find a set of network slices corresponding to that request, and slice allocation module 320 may allocate slices (e.g., selected based on a particular path cost) for service (e.g., to those haptic connections, and eventually) in response to the user request. You can.

In the illustrated example, the haptic connectivity monitoring module 330 measures a plurality of QoS values, each of which represents the QoS of a corresponding one of the plurality of haptic connections through which the respective remote controllers 125-1, 125-2, ..., 125-n are controlled. The haptic connectivity monitoring module 330 detects whether each of the plurality of haptic connections is anomalous or normal based on the measured QoS values, and if an anomalous (in other words, abnormal) among the plurality of haptic connections is detected, determines a QoS constraint based on the measured QoS values of the normal ones among the plurality of haptic connections. Admission control module 310 can then find the set of network slices that satisfy the determined QoS constraints, and slice assignment module 320 can assign a particular network slice within the set of network slices to the detected anomalous haptic connection.

Referring now to FIGS. 4-7, the operation of the network controller 155 will be described in more detail.

4 is a flow diagram showing an exemplary haptic connectivity monitoring process 400 performed by network controller 155. For example, process 400 may be performed by haptic connectivity monitoring module 330 of network controller 155 . Other example flows of process 400 are also contemplated. For example, process 400 may also include additional operations not shown and/or may include some but not all of the operations shown in FIG. 4 .

At operation 410, a plurality of haptic connections within the cooperative remote control system 100 are identified. For example, for controlling each of the cooperative remote controllers 125-1, 125-2, ..., 125-n of the cooperative remote control system 100, a haptic connection established through the network 151 between the corresponding remote controller 125 and the master device 115-1, 115-2, ..., 115-n of the cooperative remote control system 100 may be identified.

At operation 420, the QoS value of each of the plurality of haptic connections is measured. For example, the QoS value can be one or a combination of the following QoS metrics: availability, latency and delivery. The availability metric represents the ratio of service feasibility per request for a particular service. A delay metric represents the time it takes for a packet for a service to travel from the service access point to a remote target and back. The delivery metric represents the percentage of each service delivered without packet loss, i.e., the inverse of packet loss.

In some example implementations, Service Level Agreement (SLA) monitoring may be performed to measure such QoS values. SLA monitoring is the process of monitoring service health based on Network Performance Metric (NPM) measurements and QoS metrics to NPM mapping, eg, as shown in FIG. 5 . First, the NPM measurement may be performed by, for example, a network monitoring system (which may be connected to or included in the haptic connectivity monitoring module 330) by monitoring a network (e.g., network 151, and in particular, a haptic connection controlled by that device, such as remote controller 125) where a service for a given target device (e.g., remote controller 125) is provided. For example, each given NPM may be a metric related to availability, loss, delay, or utilization as shown in FIG. 5, and may be performed with a different network monitoring strategy, such as an active monitoring strategy, a passive monitoring strategy, or a Simple Network Management Protocol (SNMP) agent-based monitoring strategy. As an example, availability, loss and delay NPM may be measured according to an active monitoring scheme, which may involve generating test traffic periodically or on demand and measuring the performance of test packets or responses. After NPM measurement, a QoS metric may be measured according to a mapping function between the QoS metric and the NPM metric, such as by haptic connectivity monitoring module 330 . In a specific example, such a mapping function may be given as the following table.

Here, "device unfunctional time" represents the time when the corresponding device is not functioning, "disconnected time" represents the time when the connection with the corresponding device was disconnected, "total monitored time" represents the total time during which SLA monitoring was performed, "total number of RTT test packet" represents the total number of Round Trip Time (RTT) test packets generated for testing with the corresponding device, "ΣRTT" represents the sum of RTTs for them, and "total number of test packets" represents the test packet represents the total number of , and "number of loss packets" represents the number of lost packets among them. Referring to FIG. 5 and Table 1, availability among the QoS metrics is measured according to "device unfunctional time", "disconnected time", and "total monitored time" values related to availability among measured NPMs, delay among the measured NPMs is measured according to "total number of RTT test packet" and "ΣRTT" values related to delays among measured NPMs, and delivery among the measured NPMs is measured according to "total number of test packet" values related to delay among measured NPMs and measured NPMs. It can be seen that it is measured according to the value of "number of loss packet" related to loss.

At operation 430, an anomalous one of the plurality of haptic connections is detected. For example, such detection can be done using the QoS value measured in operation 420 .

In some example implementations, after the QoS metrics are measured, as illustrated above, whether each haptic connection is anomalous can be detected based on the following values: the current QoS value of each of the plurality of haptic connections in the cooperative remote control system 100 (QoS _t ); a previous QoS value of the corresponding haptic connection (eg, a previous QoS value obtained according to a QoS metric measurement performed immediately before this QoS metric measurement (QoS _t-1 )); and the required QoS value for that haptic connection (QoS _r ). For example, it may be determined that a haptic connection is atypical if it differs from other haptic connections in the following characteristics: the distance between the current QoS value and the desired QoS value, eg, the difference value QoS _t -QoS _r ; and/or a distance between the current QoS value and the previous QoS value, eg, a difference value ΔQoS _t =QoS _t -QoS _t-1 . Such detection would be more efficient if both of these types of difference values were used to detect the anomalous haptic connection. The former difference value indicates how far the current QoS value differs from the QoS requirements originally to be satisfied, and eventually whether the current QoS value is good. If this difference is large, it can be seen that the corresponding haptic connection is highly likely to be anomaly. In addition, assuming that the corresponding haptic connection was unusual at the time when the current QoS value was measured, but was normal at the time when the immediately preceding QoS value was measured, the latter difference value would be large. In a specific example, an atypical haptic connection may be detected by clustering each of the plurality of haptic connections in the cooperative remote control system 100 based on the aforementioned distance characteristic. Such clustering may be based on K-means clustering, Density Based Spatial Clustering of Applications with Noise, or other types of learning algorithms, such as machine learning (ML) algorithms such as decision trees, k-Nearest Neighbors (kNNs), Support Vector Machines (SVMs).

At operation 440, a request for a new network slice for the anomalous haptic connection is generated along with a QoS constraint (QoS _c ). For example, for a detected anomalous haptic connection, the QoS constraint can be set to a value that ensures that the new network slice to be provided to that haptic connection is normal, i.e. has a value somewhat similar to the QoS value of other normal haptic connections. In a specific example, the QoS constraint may be given as an average of QoS values of normal haptic connections among multiple haptic connections in the cooperative remote control system 100 . In another example, the QoS constraint may be given the minimum of the QoS values of normal haptic connections among the plurality of haptic connections.

At operation 450, a request for a new network slice is issued to the admission control mechanism for control of network 151 along with QoS constraints. For example, at operation 440 a new network slice request can be generated to include the determined QoS constraints, and then such request can be sent to an admission control mechanism, such as the admission control module 310 of the network controller 155.

For example, assume a scenario in which the cooperative remote control system 100 has three pairs of master device 115 and remote controller 125, and the three remote controllers 125 perform a common task of holding an object in the same place (hereinafter referred to simply as a "work scenario"). In the working scenario, the delay metric is used as the QoS value of the haptic connection corresponding to each pair. If one of the remote controllers 125 experiences a large delay, the task is likely to fail. In a working scenario, process 400 of FIG. 4 may be performed as follows: a haptic connection corresponding to each pair of master device 115 and remote control 125 is identified; Latency measured according to SLA monitoring; As an example, three haptic connections are currently measured to have delays of L1 _t = 0.4 ms, L2 _t = 0.6 ms, and L3 _t = 4 ms, respectively, while the previous measurements are L1 _t-1 = 0.3 ms, L2 _t-1 = 0.7 ms, and L3 _t-1 = 0.4 ms, respectively, and the required delays are all 1 ms; For each haptic connection, the distances ΔL1−, ΔL2 _t , ΔL3 _t between the current delay measure and the previous delay measure are computed, and the distances ΔL1 _rt , ΔL2 _rt , ΔL3 _rt between the current delay measure and the desired delay are computed, and using these values an anomalous haptic connection is detected (e.g., the third haptic connection is detected as an anomalous haptic connection and the remaining haptic connections are detected as normal); For a request for a new slice of an anomalous haptic connection, the QoS constraint L _c is calculated as the average of the delay measures of the normal haptic connection (eg, L _c = (L1 _t + L2 _t )/2 = 0.5 ms); and a new slice request is signaled to the admission control module 310 with the QoS constraint (L _c ) for the anomalous haptic connection.

6 is a flow diagram showing an exemplary admission control process 600 performed by network controller 155 . For example, process 600 may be performed by admission control module 310 of network controller 155 . Other example flows of process 600 are also contemplated. For example, process 600 may also include additional operations not shown and/or may include some but not all of the operations shown in FIG. 6 .

At operation 610, a new network slice request originating with the determined QoS constraints for the atypical haptic connection established on network 150 is received. At operation 620, the network slice previously provided for the anomalous haptic connection is removed. For example, removal of a network slice may involve the release of various resources such as processors, storage, bandwidth, and the like. At operation 630, information about a network slice of network 151 is obtained. For example, network slice information may be accessible from network controller 155 or read from a database included within network controller 155 . In operation 640, those of network slices in network 151 that satisfy QoS constraints are mapped to new network slice requests. At operation 650, such a slice-to-request mapping is sent to a slice allocation mechanism for control of network 151, such as slice allocation module 320 of network controller 155.

For example, in the working scenario described above with respect to FIG. 4 , in response to a new slice request R ₁ for an unusual haptic connection, the network slice serving that haptic connection may be removed. Thereafter, predetermined information about available slices of the network 151 in which the cooperative remote control system 100 operates (eg, all information regarding all QoS metrics served, such as delay, availability, delivery, etc.) can be obtained. Based on the obtained network slice information, it can be determined that, for example, two slices S ₁ and S ₂ of network 151 will guarantee a delay below the QoS constraint (L _c ) for the detected anomalous haptic connection. This mapping relationship may be signaled to the slice allocation module 320 of the network controller 155 (hereinafter, for convenience, the mapping between slice S ₁ and slice request R ₁ may be referred to as S ₁ -R ₁ , and the mapping between slice S ₂ and slice request R ₁ may be referred to as S ₂ -R ₁ , and collectively referred to as {S ₁ , S ₂ }-R ₁ ).

7 is a flow diagram showing an exemplary slice assignment process 700 performed by network controller 155 . For example, process 700 may be performed by slice allocation module 320 of network controller 155 . Other example flows of process 700 are also contemplated. For example, process 700 may also include additional operations not shown and/or may include some but not all of the operations shown in FIG. 7 .

At operation 710, a mapping between a network slice that satisfies the QoS constraints for an unusual haptic connection established on network 150 and a request for a new network slice for that haptic connection is received.

In operation 720, a path cost is calculated for each network slice mapped to the new network slice request. In some example implementations, the path cost of a network slice is the cost of creating or reconstructing that network slice. For example, any slice on the core network side may be an SFC as mentioned with respect to FIG. 2 above. SFC can be represented as a directed graph G = (V, E), where V is a set of nodes representing VMs in a slice, and E is a set of links representing connections between two VMs. Then, the path cost of a given network slice can be expressed as C=C(V)+C(E), where C is the path cost of the network slice characterized by G=(V,E), C(V) is the cost of using virtual machine nodes in the network slice, and C(E) is the cost of using connecting links between virtual machines in the network slice. As an example, the cost of using a VM node in a network slice, C(V), can be defined as the sum of the processing rates of processors such as Central Processing Units (CPUs) allocated by each VM in the network slice for the network slice. Also, as an example, the cost of using the connecting links of a network slice, C(E), may be defined as the sum of the bandwidths of each connection in the network slice.

In operation 730, the network slice with the lowest path cost is selected. In some example implementations, one may take approximate approaches such as greedy algorithms and/or use ML such as neural networks, reinforcement learning, to find the optimal network slice.

At operation 740, the selected network slice is assigned to the anomalous haptic connection.

For example, in the aforementioned working scenario, assuming that the cost of S ₁ for the S ₁ -R ₁ mapping is 45 and the cost of S ₂ for the S ₂ -R ₂ mapping is 36, then the S ₂ -R ₁ mapping (i.e. slice S ₂ for request R ₁ ) can be selected. Slice S ₂ can then be provided to serve the anomalous haptic connection.

The following are various examples related to network slice allocation in the cooperative remote control system.

In Example 1, an apparatus for network slice allocation in a collaborative teleoperation system includes the following (the cooperative teleoperation system operates on a network providing a plurality of haptic connections, but each cooperative remote controller is controlled through the plurality of haptic connections): Measuring a Quality of Service (QoS) value corresponding to each of the plurality of haptic connections, and measuring whether each of the plurality of haptic connections is unusual or normal A haptic connectivity monitoring module that detects based on the QoS value and, if an unusual one of the plurality of haptic connections is detected, determines a QoS constraint based on the measured QoS value of one or more normal ones of the plurality of haptic connections; an admission control module that identifies a set of slices of the network that satisfy the QoS constraints; and a slice allocation module that selects a particular slice in the above identified set of slices of the above network for assignment to the above anomalous haptic connection.

Example 2 includes the subject matter of Example 1, wherein measuring the QoS value includes measuring a Network Performance Metric (NPM) by performing Service Level Agreement (SLA) monitoring, and mapping the measured NPM to the QoS value.

Example 3 includes the subject matter of example 1 or example 2, wherein measuring the QoS value includes measuring the immediately preceding QoS value and the current QoS value of each of the above haptic connections, and detecting whether each of the above haptic connections is atypical or normal based on the measured QoS value determines whether each of the above haptic connections is unusual or normal for the current QoS value, the immediately preceding QoS value, and each of the above haptic connections. and detecting based on the required QoS value.

Example 4 includes the subject matter of Example 1 or Example 2, wherein measuring the QoS value includes measuring the immediately preceding QoS value and the current QoS value of each of the above haptic connections, and detecting whether each of the haptic connections is unusual or normal based on the measured QoS value determines whether each of the haptic connections is unusual or normal, a first difference between the current QoS value and the immediately preceding QoS value and the current QoS value. and detecting based on at least one of a second difference between the S value and the QoS value required for each haptic connection above.

Example 5 includes the subject matter of any of Examples 1-4, wherein the QoS constraint is an average of the measured QoS values of the one or more normal haptic connections.

Example 6 includes the subject matter of any of Examples 1-4, wherein the QoS constraint is the minimum of the measured QoS values of the one or more normal haptic connections.

Example 7 includes the subject matter of any of Examples 1-6, wherein the particular slice is to minimize path cost among the slices in the identified set.

Example 8 includes the subject matter of Example 7, wherein the particular slice is selected based on a path cost of each slice in the above identified set, any slice in the network is represented by a first set of Virtual Machines (VMs) and a second set of connections between two VMs in the first set, wherein the path cost for each slice above is given by the sum of a node usage cost and a link usage cost, wherein the node usage cost is the sum of a node usage cost and a link usage cost, wherein the node usage cost is equal to the node usage cost of each slice by each VM in the first set of each slice. is determined based on a processing rate of a processor assigned to , and the link usage cost is determined based on a bandwidth of each connection in the second set of each slice.

Example 9 includes the subject matter of any of Examples 1-8, wherein any slice of the network includes a number of ordered Virtualized Network Functions (VNFs), wherein at least one of the number of VNFs is executable VMs are provided therein.

Example 10 includes the subject matter of any of Examples 1-9, wherein selecting the particular slice includes selecting the particular slice using a greedy algorithm.

In Example 11, a method for allocating network slices in the cooperative remote control system includes the following (the cooperative remote control system operates on a network providing a plurality of haptic connections, and each cooperative remote controller is controlled through the plurality of haptic connections): Measuring a Quality of Service (QoS) value corresponding to each of the plurality of haptic connections; detecting whether each of the plurality of haptic connections is abnormal or normal based on the measured QoS value; if an unusual one of the plurality of haptic connections is detected, determining a QoS constraint based on the measured QoS value of one or more normal ones of the plurality of haptic connections; identifying a set of slices of the network that satisfy the QoS constraint; and selecting a particular slice in the above identified set of slices of the above network for assignment to the above anomalous haptic connection.

Example 12 includes the subject matter of Example 11, wherein measuring the QoS value includes measuring a Network Performance Metric (NPM) by performing Service Level Agreement (SLA) monitoring, and mapping the measured NPM to the QoS value.

Example 13 includes the subject matter of Examples 11 or 12, wherein measuring the QoS value includes measuring the immediately preceding QoS value and the current QoS value of each of the above haptic connections, and detecting whether each of the above haptic connections is anomalous or normal based on the measured QoS value determines whether each of the above haptic connections is unusual or normal. and detecting based on the QoS value required for the connection.

Example 14 includes the subject matter of Examples 11 or 12, wherein measuring the QoS value includes measuring the immediately preceding QoS value and the current QoS value of each of the haptic connections, wherein detecting whether each of the haptic connections is anomalous or normal based on the measured QoS values determines whether each of the haptic connections is unusual or normal, by determining whether each of the haptic connections is unusual or normal by performing a first difference between the current QoS value and the immediately preceding QoS value and the above current QoS value. and detecting based on at least one of a second difference between a QoS value of and a QoS value required for each haptic connection.

Example 15 includes the subject matter of any of Examples 11-14, wherein the QoS constraint is an average of the measured QoS values of the one or more normal haptic connections.

Example 16 includes the subject matter of any of Examples 11-14, wherein the QoS constraint is the minimum of the measured QoS values of the one or more normal haptic connections.

Example 17 includes the subject matter of any of Examples 11-16, wherein the particular slice is to minimize path cost among the slices in the above identified set.

Example 18 includes the subject matter of any of Examples 11-16, wherein the particular slice is selected based on a path cost of each slice in the identified set, wherein any slice of the network is represented by a first set of Virtual Machines (VMs) and a second set of connections between two VMs in the first set, wherein the path cost for each slice is given as the sum of a node usage cost and a link usage cost, wherein the node usage cost is the sum of a node usage cost and a link usage cost, wherein the node usage cost is the sum of a node usage cost and a link usage cost. determined based on the processing speed of the processor assigned to each slice by each VM in the set, and the cost of using the link is determined based on the bandwidth of each connection in the second set of each slice.

Example 19 includes the subject matter of any of Examples 11-18, wherein any slice of the network includes a number of ordered Virtualized Network Functions (VNFs) and a VM capable of running at least one of the number of VNFs is provided therein.

Example 20 includes the subject matter of any of Examples 11-19, wherein selecting the particular slice includes selecting the particular slice using a greedy algorithm.

In Example 21, a computer-readable storage medium is provided having computer-executable instructions stored thereon that, when executed by a computer processor, cause the computer processor to perform the method described in any one of Examples 11-20.

Example 22 includes the subject matter of any of Examples 1-21, wherein each of the respective cooperating remote controls includes a robot hand.

In certain instances, an apparatus, device, system, machine, etc. referred to herein may be, include, or be implemented in any suitable type of computing device. A computing device may include a processor and a computer readable storage medium readable by the processor. A processor may execute one or more instructions stored in a computer readable storage medium. The processor may also read other information stored within the computer readable storage medium. Additionally, the processor may store new information in the computer readable storage medium and may update certain information stored in the computer readable storage medium. A processor may include, for example, a central processing unit (CPU), a digital signal processor (DSP), a graphics processing unit (GPU), a processor core, a microprocessor, a microcontroller, a field-programmable gate array (FPGA), an application specific integrated circuit (ASIC), other hardware and logic circuitry, or any suitable combination thereof. . Computer readable storage media are encoded with various information, such as a set of processor executable instructions that can be executed by a processor and/or other information. For example, a computer readable storage medium may store therein computer program instructions that, when executed by a processor, cause a computing device (e.g., processor) to perform some operations disclosed herein and/or information, data, variables, constants, data structures, etc. used in those operations. Computer readable storage media include, for example, read-only memory (ROM), random-access memory (RAM), volatile memory, non-volatile memory, removable memory, non-removable memory, flash memory, solid-state memory, other types of memory devices, magnetic media such as hard disks, floppy disks and magnetic tapes, optical recordings such as CD-ROMs and DVDs. media, magneto-optical media such as floptical disks, other types of storage devices and storage media, or any suitable combination thereof.

In certain instances, any action, technique, process, or any aspect or portion thereof described herein may be embodied in a computer program product. Such computer programs may be implemented in any type of (e.g., compiled or interpreted) programming language that can be executed by a computer, such as assembly, machine language, procedural language, object-oriented language, etc., and may be combined with a hardware implementation. A computer program product may be distributed in the form of a computer readable storage medium or may be distributed online. For online distribution, some or all of the computer program product may be temporarily stored or temporarily created within a server (eg, a computer readable storage medium of a server).

The foregoing description has been presented to illustrate and describe several examples in detail. It will be appreciated by those skilled in the art that many modifications and variations can be made in light of the above teachings without departing from the scope of the present disclosure. In various instances, appropriate results may be achieved even if the techniques described above are performed in a different order and/or some of the components of the systems, architectures, devices, circuits and the like described above are combined or combined in different ways, or substituted or substituted by other components or equivalents thereof.

Therefore, the scope of the present disclosure should not be limited to the form disclosed, and should be defined by the claims described below and equivalents thereof.

100: cooperative remote control system
110: master domain
120: slave domain
150: network domain

Claims (15)

An apparatus for network slice allocation in a collaborative teleoperation system, wherein the cooperative teleoperation system operates on a network providing a plurality of haptic connections, wherein respective cooperative remote controllers are controlled through the plurality of haptic connections, the apparatus is disposed on a core side of the network, the apparatus comprising:
Measuring a Quality of Service (QoS) value corresponding to each of the plurality of haptic connections, detecting whether each of the plurality of haptic connections is abnormal or normal based on the measured QoS value, and if an unusual one of the plurality of haptic connections is detected, determining a QoS constraint based on the measured QoS value of one or more normal ones of the plurality of haptic connections, and signaling a new slice request including the determined QoS constraint a haptic connectivity monitoring module;
an admission control module that, in response to the new slice request, identifies a set of slices of the network that satisfy the QoS constraint;
a slice assignment module that selects a particular slice in the identified set of slices of the network for assignment to the atypical haptic connection.
Device. According to claim 1,
Measuring the QoS value includes measuring a Network Performance Metric (NPM) by performing Service Level Agreement (SLA) monitoring, and mapping the measured NPM to the QoS value,
Device. According to claim 1,
Measuring the QoS value includes measuring a previous QoS value and a current QoS value of each haptic connection, and detecting whether each haptic connection is abnormal or normal based on the measured QoS value determines whether each haptic connection is unusual or normal A first difference between the current QoS value and the previous QoS value and a QoS value required for each haptic connection and the current QoS value Including detecting based on at least one of the second differences between
Device. According to claim 1,
wherein the QoS constraint is an average of the measured QoS values of the one or more normal haptic connections;
Device. According to claim 1,
Wherein the QoS constraint is the minimum of the measured QoS values of the one or more normal haptic connections;
Device. According to claim 1,
The particular slice is selected based on the path cost of each slice in the identified set, and any slice of the network is represented by a first set of Virtual Machines (VMs) and a second set of connections between two VMs in the first set, for each slice, the path cost is given by the sum of a node usage cost and a link usage cost, wherein the node usage cost is the processing rate of a processor assigned to each slice by each VM in the first set of each slice. wherein the link usage cost is determined based on the bandwidth of each connection in the second set of each slice.
Device. According to claim 1,
Selecting the particular slice comprises selecting the particular slice using a greedy algorithm.
Device. A method performed by a network controller for network slice allocation in a cooperative remote control system, wherein the cooperative remote control system operates on a network providing a plurality of haptic connections, wherein each cooperative remote controller is controlled through the plurality of haptic connections, the network controller is disposed on a core side of the network, the method comprising:
measuring a Quality of Service (QoS) value corresponding to each of the plurality of haptic connections;
detecting whether each of the plurality of haptic connections is abnormal or normal based on the measured QoS value;
if an anomalous one of the plurality of haptic connections is detected, determining a QoS constraint based on the measured QoS value of at least one normal one of the plurality of haptic connections and signaling a new slice request including the determined QoS constraint;
in response to the new slice request, identifying a set of slices of the network that satisfy the QoS constraint;
selecting a particular slice in the identified set of slices of the network for assignment to the atypical haptic connection.
method. According to claim 8,
The step of measuring the QoS value includes measuring a Network Performance Metric (NPM) by performing Service Level Agreement (SLA) monitoring, and mapping the measured NPM to the QoS value,
method. According to claim 8,
The step of measuring the QoS value includes measuring the previous QoS value and the current QoS value of each haptic connection, and the step of detecting whether each haptic connection is abnormal or normal based on the measured QoS value determines whether the respective haptic connection is unusual or normal, a first difference between the current QoS value and the previous QoS value and the current QoS value and a required Q for each haptic connection. Including detecting based on at least one of the second differences between the oS values,
method. According to claim 8,
wherein the QoS constraint is an average of the measured QoS values of the one or more normal haptic connections;
method. According to claim 8,
Wherein the QoS constraint is the minimum of the measured QoS values of the one or more normal haptic connections;
method. According to claim 8,
the particular slice is selected based on the path cost of each slice in the identified set, and any slice in the network is represented by a first set of Virtual Machines (VMs) and a second set of connections between two VMs in the first set, for each slice, the path cost is given by the sum of a node usage cost and a link usage cost, the node usage cost being determined by each VM in the first set of each slice based on the processing speed of the processor assigned to the respective slice; wherein the link usage cost is determined based on the bandwidth of each connection in the second set of each slice.
method. According to claim 8,
Wherein selecting the particular slice comprises selecting the particular slice using a greedy algorithm.
method. A computer-readable storage medium having stored thereon computer-executable instructions which, when executed by a computer processor, cause the computer processor to perform a method according to any one of claims 8 to 14.