Computer systems with hierarchical temporal memory (HTM)

Computer systems with hierarchical temporal memory (HTM)

Background and purpose:

Access control: Access control is a method for managing access to restricted resources by users or services in computer systems, and it is necessary to define access rights rules for users or consumers.

Existing problems: Defining access control rules requires a lot of effort, and the rules may become complicated due to multi-dimensional considerations.

Solution: A computer implementation method based on HTM is proposed to control access to restricted resources, aiming to reduce the burden of defining access control rules.

HTM access control method:

Training phase: Receive the first set of records of the computer system during the training time, which describes in detail the authorized access to the resource. Generate a sparse distributed representation (SDR) of these records, and train the HTM based on the SDR to build an operation model of the computer system during the training period.

Operation phase: Receive the second set of records of the system during the operation time, and also generate SDR. Execute the trained HTM, identify the input set records based on the model, and if the recognition degree is lower than the threshold, it is determined to be an unauthorized use of the resource by the consumer.

HTM technical principle:

HTM model: HTM is a neural network model that mimics the function of the cerebral cortex. It is used to process time series data and learn and predict by identifying patterns and sequence relationships in the input.

SDR representation: SDR is a data encoding method that distributes information in a set of binary vectors, with only a few bits activated. This representation method helps HTM to efficiently process and identify data.

Method steps:

Receive the first set of records: During training, collect system records about authorized access to resources.

Generate SDR and train HTM: Convert records to SDR, and use these SDRs to train HTM and build an operation model.

Receive the second set of records: During operation, collect system records about resource usage.

Execute HTM and identify: Convert new records to SDR, execute HTM to identify the consistency of records with the model, and detect unauthorized use.

Blockchain combined with access control:

Blockchain application: In some embodiments, blockchain technology is used to allocate and manage digital currency (such as Bitcoin) when consumers access resources, and a consumable access control mechanism is used to ensure compliance with resource access.

Access control enhancement: When HTM detects unauthorized use, the digital currency allocated to the consumer is consumed through blockchain transactions until further restrictive measures are taken after it is exhausted.

System architecture and components:

System components: including computer systems, restricted resources, access control systems, HTM training and execution components, etc.

Interaction process: Access control is automated and intelligent through steps such as logging, SDR generation, HTM training and execution between components.

Implementation and effect:

Specific implementation: The implementation process of the HTM access control method is demonstrated through detailed component diagrams and flow charts, including specific steps in the training phase and the operation phase.

Expected effect: This method can efficiently and accurately identify unauthorized use of restricted resources, and enhance the reliability and security of access control through blockchain technology.

What is HTM and what is its role in access control?

Answer: HTM (Hierarchical Temporal Memory) is a neural network model that mimics the function of the cerebral cortex and is used to process time series data. In access control, HTM identifies authorized and unauthorized use of restricted resources by building an operational model of the computer system during training. Specifically, HTM receives system records during training time and operation time, generates training sets and input sets through sparse distributed representation (SDR), thereby learning and identifying normal operation modes, and triggering the determination of unauthorized use when anomalies are detected.

What are the main stages of the access control method proposed in? What are the main tasks of each stage?

Answer: The access control method proposed in mainly includes a training stage and an operation stage. In the training stage, the main task is to receive records from the computer system during the training time, generate SDR representations of these records, and train HTM based on SDR to build an operation model. In the operation stage, the main task is to receive records from the system during the operation time, generate SDR representations, execute the trained HTM to identify the consistency between the records and the model, and determine whether it is an unauthorized use of resources by consumers based on the recognition degree.

Briefly describe the role of SDR in the HTM access control method.

Answer: SDR (Sparse Distributed Representation) plays a key role in the HTM access control method. It is an efficient data encoding method that distributes information in a set of binary vectors with only a few bits activated. SDR representation helps HTM process and identify data efficiently because it can capture key features in input data while reducing redundant information. In the training phase, SDR is used to generate the training set of HTM; in the operation phase, SDR is used to generate the input set of HTM, thereby realizing the identification and analysis of new records.

How is blockchain technology combined with HTM access control methods?

Answer: The combination of blockchain technology and HTM access control methods is mainly reflected in the allocation and management of digital currency when accessing resources. In some embodiments, digital currency (such as Bitcoin) is allocated to consumers through blockchain technology to ensure the compliance of resource access through a consumption-based access control mechanism. When HTM detects unauthorized use of resources by consumers, the digital currency allocated to consumers is consumed through blockchain transactions. Once the digital currency is exhausted, the system will take further restrictive measures, such as prohibiting consumers from accessing resources. This combination enhances the reliability and security of access control because blockchain technology provides decentralized, transparent and tamper-proof transaction records.

Please summarize the main advantages of the HTM access control method proposed in.

A: The main advantages of the HTM access control method proposed in the paper include: efficiency, HTM can process time series data and quickly identify abnormal operations; accuracy, by building precise operation models to determine unauthorized use; flexibility, it can adapt to the access control requirements of different scenarios and resource types; and security, combined with blockchain technology to enhance the reliability and transparency of access control. In addition, this method also reduces the burden of defining complex access control rules and improves management efficiency.