UNIT – 4

Mobile Agents in Computing

Mobile agents are software entities that can move autonomously across different network
environments to perform tasks on behalf of a user or another program. They are designed to
operate in distributed systems and can carry code, data, and state information with them as
they migrate from one host to another. Mobile agents offer several advantages in terms of
flexibility, efficiency, and scalability for various applications in mobile and distributed
computing.
Key Concepts:
1. Autonomy:
 Mobile agents operate independently without requiring continuous interaction
with a central control system. They make decisions and perform tasks
autonomously based on the logic encoded within them.
2. Mobility:
 The defining feature of mobile agents is their ability to move from one host to
another within a network. This movement can be triggered by the agent itself or
by external conditions specified by its code.
3. Interactivity:
 Mobile agents can interact with other agents, systems, and resources at the
remote host. They can gather information, perform computations, and modify
data.
4. Adaptability:
 Mobile agents can adapt their behavior based on the environment they
encounter at different hosts. This enables them to efficiently handle dynamic and
heterogeneous network conditions.
5. State Preservation:
 Mobile agents carry their state information with them, allowing them to resume
operations seamlessly at the new host. This includes variables, execution state,
and collected data.
Applications of Mobile Agents:
1. Distributed Information Retrieval:
 Mobile agents can travel across various information sources to gather and
aggregate data, reducing the amount of data transferred over the network and
minimizing latency.
2. Network Management:
 Mobile agents can be used to monitor and manage network resources, such as
configuring network devices, diagnosing faults, and optimizing traffic flow.
 3. E-commerce:
 In e-commerce, mobile agents can represent buyers or sellers, autonomously
negotiating deals, comparing prices, and finalizing transactions.
4. Load Balancing:
 Mobile agents can be deployed to analyze and redistribute workloads across
servers in a distributed system, enhancing performance and resource utilization.
5. Remote Diagnostics and Maintenance:
 Mobile agents can be sent to remote locations to perform diagnostics, update
software, or carry out maintenance tasks on devices and systems.
Advantages of Mobile Agents:
1. Reduced Network Load:
 By moving computations closer to the data, mobile agents can reduce the
amount of data that needs to be transmitted over the network.
2. Asynchronous Execution:
 Mobile agents operate asynchronously, meaning they can continue functioning
even when disconnected from their originating host.
3. Scalability:
 Mobile agents can dynamically scale by distributing themselves across multiple
hosts, handling increasing loads efficiently.
4. Fault Tolerance:
 Mobile agents enhance fault tolerance by being able to migrate away from failing
hosts and continue operations elsewhere.
Challenges and Issues:
1. Security:
 Security is a significant concern for mobile agents. They must be protected
against malicious hosts, and hosts must be safeguarded against malicious agents.
Encryption, authentication, and sandboxing techniques are commonly employed
to address these concerns.
2. Standardization:
 Lack of standardization in mobile agent platforms and protocols can hinder
interoperability and widespread adoption.
3. Resource Management:
 Managing resources, such as CPU, memory, and network bandwidth, across
various hosts can be complex, especially when agents operate in heterogeneous
environments.
4. Fault Tolerance:
  Ensuring the reliability and fault tolerance of mobile agents in the face of
network disruptions or host failures is challenging and requires robust
mechanisms for state preservation and recovery.
5. Performance Overhead:
 The process of migrating agents between hosts can introduce performance
overhead due to serialization, transmission, and deserialization of agent state
and code.
Technologies and Platforms:
1. Java Aglets:
 Aglets are Java-based mobile agents that use the Aglet API to facilitate the
creation, migration, and communication of mobile agents across networked
hosts.
2. IBM's Aglet Workbench:
 This platform provides a comprehensive framework for developing and
deploying mobile agents using Java Aglets.
3. Telescript:
 An early mobile agent language developed by General Magic, Telescript
facilitated the creation and execution of mobile agents on various networked
hosts.
4. Jade (Java Agent DEvelopment Framework):
 JADE is a middleware platform that supports the development of multi-agent
systems, including mobile agents, using the Java programming language.

Security and Fault Tolerance in Mobile Agent Computing

Mobile agents, while offering numerous advantages in distributed computing, face significant
challenges in terms of security and fault tolerance. Addressing these issues is crucial to ensuring
the reliability and safety of mobile agent-based systems.
Security in Mobile Agent Computing
1. Security Threats:
 Malicious Hosts: Hosts that can potentially modify, delete, or misuse the mobile agents.
 Malicious Agents: Agents that can attack the host or other agents, steal data, or disrupt
services.
 Eavesdropping: Intercepting communication between agents and hosts to gain
unauthorized access to sensitive information.
 Masquerading: Impersonating legitimate agents or hosts to gain unauthorized access
or perform malicious actions.
2. Security Mechanisms:
a. Authentication:
 Ensures that agents and hosts can verify each other's identities.
 Techniques include digital signatures and certificates.
b. Encryption:
 Protects the confidentiality and integrity of data transmitted between agents and hosts.
 Symmetric and asymmetric encryption algorithms can be used.
c. Access Control:
 Determines the permissions granted to agents based on their identities and roles.
 Access control lists (ACLs) and role-based access control (RBAC) are common
approaches.
d. Sandboxing:
 Isolates the execution environment of agents to prevent them from accessing
unauthorized resources or causing harm.
 The Java Virtual Machine (JVM) is an example of a sandboxing environment.
e. Code Signing:
 Ensures that the agent's code has not been tampered with by verifying its digital
signature.
 Provides authenticity and integrity guarantees for the agent's code.
f. Secure Migration Protocols:
 Ensure secure transfer of agents between hosts, protecting against interception and
tampering.
 Techniques like SSL/TLS can be used to secure migration channels.
Fault Tolerance in Mobile Agent Computing
1. Fault Tolerance Challenges:
 Agent Failure: Agents might fail due to software bugs, resource exhaustion, or hostile
environments.
 Host Failure: The host running the agent might crash or become unavailable.
 Network Failure: Network partitions or interruptions can disrupt the communication
between agents and hosts.
2. Fault Tolerance Mechanisms:
a. Checkpointing:
 Periodically saves the state of the mobile agent to stable storage.
 Upon failure, the agent can be restarted from the last checkpoint.
b. Replication:
  Creates multiple copies of the agent to run on different hosts.
 If one host fails, other copies can continue the operation.
c. Migration Strategies:
 Proactive Migration: Moves agents to a new host before a potential failure (e.g., based
on load or health monitoring).
 Reactive Migration: Moves agents to a new host after detecting a failure.
d. Redundancy:
 Use of redundant systems and paths to ensure availability in case of component failures.
e. Agent Collaboration:
 Agents can collaborate and share their state or results with peer agents to mitigate the
impact of individual failures.
f. Monitoring and Self-Healing:
 Continuous monitoring of agent and host health.
 Automated recovery actions, such as restarting failed agents or reallocating tasks to
healthy agents.
Implementation Examples:
1. Secure Agent Platforms:
 Platforms like Java Aglets and Jade offer built-in security features such as authentication,
encryption, and sandboxing.
2. Fault-Tolerant Middleware:
 Middleware solutions that provide checkpointing, replication, and migration capabilities
to enhance fault tolerance.

Transaction Processing in Mobile Computing Environments

Transaction processing in mobile computing environments involves ensuring that all
transactions (e.g., financial transactions, database updates, or any series of operations) are
executed in a reliable, consistent, and secure manner, even when the network connectivity is
intermittent, resources are constrained, and the mobile devices are constantly moving. Here’s
an in-depth look at the concepts, challenges, and mechanisms involved in transaction processing
in mobile computing:
Key Concepts:
1. Transaction Properties:
 ACID Properties: Transactions should maintain the ACID properties—
Atomicity, Consistency, Isolation, and Durability.
 Atomicity: Ensures that all operations within a transaction are
completed successfully. If any operation fails, the entire transaction is
rolled back.
  Consistency: Ensures that a transaction transforms the system from one
valid state to another valid state.
 Isolation: Ensures that concurrent transactions do not interfere with
each other.
 Durability: Ensures that once a transaction is committed, it remains so,
even in the case of system failures.
2. Mobile Transaction Models:
 Kangaroo Transactions: These involve a primary transaction that can hop
between various base stations (hosts) and sub-transactions that are executed at
different locations.
 Chained Transactions: Transactions are broken into smaller sub-transactions
that are executed sequentially across different mobile hosts.
 Pro-Motion Transactions: Incorporate pre-transaction and post-transaction
phases to handle the mobile environment's disconnections and reconnections.
Challenges in Mobile Transaction Processing:
1. Intermittent Connectivity:
 Mobile devices often experience intermittent or unreliable network connections,
which can disrupt transaction processing.
2. Limited Resources:
 Mobile devices typically have limited computational power, storage, and battery
life compared to stationary systems.
3. Handoff Issues:
 When a mobile device moves between different network cells or access points,
maintaining the continuity and integrity of transactions can be challenging.
4. Concurrency Control:
 Managing concurrent transactions in a mobile environment where multiple
devices may access and modify the same data.
5. Consistency Maintenance:
 Ensuring that the database remains consistent despite frequent disconnections
and partial transactions.
Mechanisms and Strategies:
1. Caching and Prefetching:
 Data Caching: Mobile devices cache frequently accessed data locally to reduce
dependency on continuous network connectivity.
 Prefetching: Predictive techniques are used to preload data that the mobile
device might need, reducing latency and dependency on the network.
2. Checkpointing and Logging:
  Checkpointing: Periodically saving the state of the transaction so that it can be
resumed from the last checkpoint in case of failures.
 Logging: Maintaining logs of all transaction operations to ensure that
transactions can be rolled back or replayed if needed.
3. Asynchronous Processing:
 Allowing transactions to proceed independently of network availability, with
changes being synchronized once connectivity is restored.
4. Two-Phase Commit (2PC):
 A distributed algorithm that ensures all participants in a transaction agree on
the transaction's outcome (commit or abort), even in the presence of failures.
5. Optimistic Concurrency Control:
 Assumes that conflicts are rare and allows transactions to execute without
restrictions initially. Conflicts are checked at commit time, and if a conflict is
detected, the transaction is rolled back.
6. Compensating Transactions:
 Used to undo the effects of a partially completed transaction. If a transaction
cannot be completed, compensating transactions ensure that the system returns
to a consistent state.
Examples of Transaction Processing Systems:
1. Mobile Banking Applications:
 Handle financial transactions securely and reliably, even with intermittent
connectivity. Techniques like secure offline caching and deferred
synchronization are often used.
2. Mobile Commerce (m-commerce):
 Shopping apps that process orders, payments, and inventory updates. These
systems use transaction models that can handle disconnections and ensure
consistency across different systems.
3. Distributed Database Systems:
 Enable mobile devices to participate in distributed transactions by providing
mechanisms for local transaction management and synchronization with the
central database when connectivity is restored.