Abstract: A method for generating, conveying, and using attempted producer network function (NF) instance communication information includes, at a first service communication proxy (SCP), receiving, from a sender, a first service based interface (SBI) request message. The method further includes attempting to obtain a service requested by the first SBI request message from at least one producer NF instance. The method further includes receiving at least one error response or failing to receive a response from the at least one producer NF instance. The method further includes generating, from the at least one error response or the failing to receive a response from the at least one producer NF instance, attempted producer NF instance communication information. The method further includes communicating, to the sender, the attempted producer NF instance communication information.

Abstract: A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate for multiplying the stored bit with an input vector bit. An output node for the logic gate connects to a second plate of a capacitor. A first plate of the capacitor connects to a read bit line. A write driver controls a power supply voltage to the cross-coupled inverters, the first switch, and the second switch to capacitively write the stored bit to the pair of cross-coupled inverters.

Abstract: Certain aspects of the present disclosure provide techniques for efficient depthwise convolution. A convolution is performed with a compute-in-memory (CIM) array to generate CIM output, and at least a portion of the CIM output corresponding to a first output data channel, of a plurality of output data channels in the CIM output, is written to a digital multiply-accumulate (DMAC) activation buffer. A patch of the CIM output is read from the DMAC activation buffer, and weight data is read from a DMAC weight buffer. Multiply-accumulate (MAC) operations are performed with the patch of CIM output and the weight data to generate a DMAC output.

Abstract: A network system can match a service provider with multiple service requests. The network system can further determine a service plan that indicate an ordered sequence of service locations for the service provider in fulfilling the multiple service requests. A route can be determined for the service provider based on the service plan. While the service provider is in-progress of fulfilling the multiple service requests, the network system can detect a service plan update event. In response to detecting the service plan update event, the network system can update the service plan of the service provider, including determining a different order of service locations for the service provider. An updated route can be generated for the service provider based on the updated service plan.

Abstract: A method for distributing network function (NF) high availability (HA) topology information in a core network includes, at an NF repository function (NRF) including at least one processor, receiving, from a plurality of producer NFs in an NF set, NFRegister requests including NF HA topology information for the producer NFs. The method further includes registering the producer NFs and storing the NF HA topology information for the producer NFs. The method further includes receiving, from a consumer NF or service communication proxy (SCP), an NFDiscover request containing at least one service discovery parameter that corresponds to a service provided by the producer NFs. The method further includes responding to the NFDiscover request by generating an NFDiscover response, including, in the NFDiscover response, the NF HA topology information for the producer NFs, and transmitting the NFDiscover response to the consumer NF or SCP.

Abstract: The subject matter described herein includes a method for reducing the likelihood of successful denial of service (DoS) attacks by validating overload control information (OCI) scope information against network function (NF) profile information obtained using target resource identification information. The method includes receiving a service based interface (SBI) request message, obtaining, from the SBI request message, target resource identification information, obtaining NF profile information using the target resource identification information and storing the NF profile information, receiving an SBI response message including overload control information and scope information for the overload control information, using the stored NF profile information to determine whether the scope information for the overload control information is valid, and, in response to determining that the scope information for the overload control information is invalid, rejecting the SBI response message.

Abstract: The present disclosure relates to systems, methods, and computer-readable media for receiving usage data for a virtual device (or other virtual service), analyzing the usage data to determine a usage bucket characteristic of usage of the virtual device over a period of time, and determining a usage score for the virtual device. The systems described herein further involve causing a deployment of the virtual device to be upgraded, downgraded, or otherwise modified based on the usage bucket and associated usage score. The features and functionalities described herein can provide an efficient mechanism for administrating a tenant deployment as well as implementing a more efficient utilization of cloud computing resources for a variety of virtual services.

Abstract: A method for optimized routing of service based interface (SBI) request messages to remote network function (NF) repository functions (NRFs) using indirect communications via a service communications proxy (SCP) includes, at an SCP including at least one processor, receiving an SBI request message. The method further includes forwarding the SBI request message to a remote NRF. The method further includes determining that the remote NRF is unable to process the SBI request message, and, in response to determining that the remote NRF is unable, identifying a georedundant mate of the remote NRF. The method further includes forwarding the SBI request message to the georedundant mate NRF of the remote NRF that is unable to process the SBI request message.

Abstract: A near memory system is provided for the calculation of a layer in a machine learning application. The near memory system includes an array of memory cells for storing an array of filter weights. A multiply-and-accumulate circuit couples to columns of the array to form the calculation of the layer.

Abstract: Methods, systems, and computer readable media for rank processing in network function selection. A method includes periodically receiving, at a network function discovery node, and from each producer network function of a number of producer network functions, a current load value specifying a computing load carried by the producer network function. The network function discovery node is configured for performing service discovery between network functions of a telecommunications core network. The method includes determining, for each producer network function, an available capacity for the producer network function based on the current load value and a published capacity of the producer network function. The method includes responding to a network function discovery request from a consumer network function using the available capacity of each producer network function.

Abstract: A method for generating and using network function (NF) set load information, the method includes, at a service communications proxy (SCP), receiving service based interface (SBI) requests from consumer NFs. The method further includes forwarding the SBI requests to producer NF instances that are members of an NF set. The method further includes receiving responses to the SBI requests from the producer NF instances. The method further includes determining NF instance load control information (LCI) for the producer NF instances using the responses. The method further includes computing, by the SCP and from the NF instance LCI for the producer NF instances, NF set LCI for the NF set. The method further includes communicating the NF set LCI for the NF set to at least one of the consumer NFs or using the NF set LCI for the NF set to select a producer NF instance within an NF set to provide a service for one of the consumer NFs.

Abstract: A method for generating network function (NF) set load information aware NF discovery responses includes, at an NF repository function (NRF), receiving NFUpdate messages from producer NF instances. The method further includes collecting or generating, from the NFUpdate messages, NF set load information for NF sets of which the producer NF instances are members. The method further includes receiving, from a consumer NF, an NF discovery request. The method further includes generating, using query parameters in the NF discovery request, an NF discovery response including NF profiles of producer NF instances corresponding to the query and, including, in the NF discovery response, NF set load information for NF sets of producer NF instances whose NF profiles are included in the NF discovery response. The method further includes forwarding the NF discovery response including the NF set load information to the consumer NF.

Abstract: A method for resource object level authorization at a network function (NF) includes maintaining, by a first NF, a service based interface (SBI) resource object access authorization policy database containing policies for controlling access to SBI resource objects and dynamically populating a resource object owner database containing records for resource objects and corresponding resource object owners. The method further includes receiving, by the first NF and from a second NF, a first SBI resource object access request for accessing a resource object, accessing, using information from the first SBI resource object access request, the resource object access authorization policy database and the resource object owner database, determining that an access to the resource object requested by the first resource object access request is not permitted, and preventing the access to the resource object requested by the first resource object access request.

Abstract: A method for optimizing network traffic distribution using timeslot-based tracked producer network function (NF) performance during producer NF selection is performed at a service communications proxy (SCP) including at least one processor. The method includes tracking a performance characteristic of producer NFs across different timeslots. The method further includes receiving, from a consumer NF, a service based interface (SBI) service request. The method further includes identifying eligible producer NFs capable of handling the SBI service request. The method further includes selecting, from the eligible producer NFs, in a manner that at least partially overrides priorities of the eligible producer NFs, and based on the tracked performance characteristic across the different timeslots, a producer NF to handle the SBI service request. The method further includes forwarding the SBI service request to the producer NF.

Abstract: A method for resolution of inter-network domain names between telecommunications networks includes storing, at a security edge protection proxy (SEPP) of a home network, a mapping between a domain name and a network address of a producer network function of the home network. The method includes receiving, at the SEPP of the home network, a request message from a consumer network function of a visitor network. The method includes resolving, at the SEPP of the home network, a request message domain name of the request message using the mapping between the domain name and the network address of the producer network function of the home network.

Abstract: A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate formed by a pair of switches for multiplying the stored bit with an input vector bit. A controller controls the pair of switches responsive to a sign bit during a computation phase of operation and controls the pair of switches responsive to a magnitude bit during an execution phase of operation.

Abstract: A secure enclave may be used to satisfy privacy requirements and audit requirements. Code may be loaded into the secure enclave. The code may generate a predefined report based on data and added noise. The pre-defined report may be subject to audit requirements. The data may be subject to the privacy requirements. The secure enclave may generate an encryption key and a decryption key based on the code. Only the secure enclave may have access to the decryption key. And the secure enclave may allow only a verified copy of the code to access the decryption key. With the added noise, the report may satisfy a pre-defined differential privacy guarantee. Encrypting the code and ensuring that the report satisfies the differential privacy guarantee may satisfy the privacy requirements. Retaining the report, the code, the secure enclave, and the encrypted data may satisfy the audit requirements.

Abstract: A method for automatically managing a platform firewall using a network function (NF) repository function (NRF) or service communications proxy (SCP) includes receiving message relating to registering, updating or deregistering an NF profile in an NF profiles database separate from a platform firewall. The method further includes determining that the registering, updating, or deregistering of the NF profile requires a change to a firewall rules configuration of the platform firewall. The method further includes, in response to determining that the registering, updating, or deregistering of the NF profile requires a change to the firewall rules configuration of the platform firewall, automatically updating, by the NRF or SCP, the firewall rules configuration of the platform firewall.

Abstract: A method for distributing network function (NF) high availability (HA) topology information in a core network includes, at an NF repository function (NRF) including at least one processor, receiving, from a plurality of producer NFs in an NF set, NFRegister requests including NF HA topology information for the producer NFs. The method further includes registering the producer NFs and storing the NF HA topology information for the producer NFs. The method further includes receiving, from a consumer NF or service communication proxy (SCP), an NFDiscover request containing at least one service discovery parameter that corresponds to a service provided by the producer NFs. The method further includes responding to the NFDiscover request by generating an NFDiscover response, including, in the NFDiscover response, the NF HA topology information for the producer NFs, and transmitting the NFDiscover response to the consumer NF or SCP.

Abstract: An apparatus for electrically and mechanically coupling battery modules to an electric vehicle includes an interface plate and one or more battery trays. The interface plate includes an electrical output that electrically couples the interface plate to the vehicle. The interface plate also includes an electrical input that is electrically coupled to the electrical output. The battery tray is configured to receive a plurality of replaceable battery modules and is releasably mechanically coupled to the interface plate.