Introduction to multi-agent modified Q-learning routing for computer networks

Rudek, Radosław; Koszalka, Leszek; Pozniak-Koszalka, Iwona

doi:10.1109/aict.2005.53

Cited by 6 publications

(8 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we will show that scheduling problems with the learning effect (see [8] or [11]) allow us to model such problems as a minimization of the total transmission cost in a computer network that uses Modified Q-learning Routing algorithm (see [30]), i.e., a reinforcement learning method.…”

Section: Scheduling Problem Formulationmentioning

confidence: 99%

“…The parameters of the network are as follows: (i) link cost: 100; (ii) packet size: 5 flits; (iii) number of packets generated per simulation step: 75; (iv) simulation steps: 400. The parameters of the MQR algorithm are given: α = 0.7, γ = 0.9 with switched off forgetting factor, i.e., φ = 1 and φ p =0 (see [30]). It can be seen (Figure 4) that the handling of packets by MQR network according to HA NI is characterized by the lower cost than the network with a random sequencing or than the cost provided by with HA ND algorithm.…”

Section: Propertymentioning

confidence: 99%

“…In each network node operates MQR agent [30] that cooperates with other agents to work out the globally minimal total cost of transmissions. Since, we consider mesh topology of the network, let us denote each node (i.e., agent) by (x, y), where x is the row and y is the column in the mesh (x, y ∈ {1, .…”

Section: Scheduling Problem Formulationmentioning

confidence: 99%

See 2 more Smart Citations

On Scheduling Problems with an Intelligent Use of the Learning Effect

Janiak

Rudek

Janiak

2009

2009 First Asian Conference on Intelligent Information and Database Systems

View full text Add to dashboard Cite

This paper is devoted to scheduling problems with the learning effect, which is understood as a process of acquiring experience that increases the efficiency of a processor. To bring closer the considered phenomenon, a short survey on results concerning scheduling problems with the learning effect is provided. In particular, the existing models of the experience are presented along with a discussion on different shapes of the learning curve. Some complexity results of scheduling problems with the learning effect are also presented. We also show that scheduling problems with the learning effect model such problems as a minimization of a total transmission cost of packets in a computer network that uses a reinforcement learning routing algorithm. We also derive properties that allow us to construct scheduling algorithms, which can be applied in the computer network to increase its effectiveness by the utilization of its learning ability.

show abstract

Section: Scheduling Problem Formulationmentioning

confidence: 99%

Section: Propertymentioning

confidence: 99%

See 1 more Smart Citation

On Scheduling Problems with an Intelligent Use of the Learning Effect

Janiak

Rudek

Janiak

2009

2009 First Asian Conference on Intelligent Information and Database Systems

View full text Add to dashboard Cite

show abstract

“…They have been used in robotics, to study environments with multiple autonomous robots [1,14,17,23], and control systems to study motion of mobile robots [22,26]. They have also been used in telecommunications, in conjunction with Q-Learning [2,19,20], to enhance routing techniques [18], in power engineering, exploring the "decisions-inherent in engineering multi-agent systems" for power-related applications [12,13], and in power systems, using multiagent reinforcement learning to solve the problems that arise from the nonlinearity of a power system [4].…”

Section: Introductionmentioning

confidence: 99%

“…Multiagent systems have been well studied in the literature and used to solve complex problems relying on the value of collaboration among multiple intelligent agents or expert systems [6,8,11,18,25]. They have been used in robotics, to study environments with multiple autonomous robots [1,14,17,23], and control systems to study motion of mobile robots [22,26].…”

Section: Introductionmentioning

confidence: 99%

Reinforcement learning technique using agent state occurrence frequency with analysis of knowledge sharing on the agent’s learning process in multiagent environments

Al-Dayaa

Megherbi

2010

J Supercomput

View full text Add to dashboard Cite

Reinforcement learning techniques like the Q-Learning one as well as the Multiple-Lookahead-Levels one that we introduced in our prior work require the agent to complete an initial exploratory path followed by as many hypothetical and physical paths as necessary to find the optimal path to the goal. This paper introduces a reinforcement learning technique that uses a distance measure to the goal as a primary gauge for an autonomous agent's action selection. In this paper, we take advantage of the first random walk to acquire initial information about the goal. Once the agent's goal is reached, the agent's first perceived internal model of the environment is updated to reflect and include said goal. This is done by the agent tracing back its steps to its origin starting point. We show in this paper, no exploratory or hypothetical paths are required after the goal is initially reached or detected, and the agent requires a maximum of two physical paths to find the optimal path to the goal. The agent's state occurrence frequency is introduced as well and used to support the proposed Distance-Only technique. A computation speed performance analysis is carried out, and the Distance-and-Frequency technique is shown to require less computation time than the Q-Learning one. Furthermore, we present and demonstrate how multiple agents using the Distance-and-Frequency technique can share knowledge of the environment and study the effect of that knowledge sharing on the agents' learning process.

show abstract

An Intelligent Mechanism for Utility and Active Customers in Demand Response Using Single and Double Q Learning Approach

Chandrakar¹,

Paliwal²

2022

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

Introduction to multi-agent modified Q-learning routing for computer networks

Cited by 6 publications

References 6 publications

On Scheduling Problems with an Intelligent Use of the Learning Effect

On Scheduling Problems with an Intelligent Use of the Learning Effect

Reinforcement learning technique using agent state occurrence frequency with analysis of knowledge sharing on the agent’s learning process in multiagent environments

An Intelligent Mechanism for Utility and Active Customers in Demand Response Using Single and Double Q Learning Approach

Contact Info

Product

Resources

About