Page 1 of 12

Transactions on Engineering and Computing Sciences - Vol. 13, No. 1

Publication Date: February 25, 2025

DOI:10.14738/tecs.131.18272.

Huang, M.-C. (2025). Apply Reputation Proof with Load Services in Ad-hoc Networks. Transactions on Engineering and Computing

Sciences, 13(1). 87-98.

Services for Science and Education – United Kingdom

Apply Reputation Proof with Load Services in Ad-hoc Networks

Ming-Chang Huang

Department of Business Information Systems / Operation

Management University of North Carolina at Charlotte

ABSTRACT

In this paper, a new system design for load services in computer networks with a

new reputation system is constructed to check available host reputation. Database

systems are used for directory agents to save information provided by load-server

agents and protocols are built how a host can find available hosts for load service

and load transfer purposes when the host moves to a new region. This includes how

a directory agent builds its database, how a load-server agent provides its services,

and how a load-client agent gets its desired services. The system uses the fuzzy logic

control method to transfer loads for load balancing, instead of fixed threshold level

methods. The purpose of this new system structure is to provide efficient ways in

building communication and accessing resources in ad-hoc computer network

systems and helps users to find resources easily and securely.

Keywords: Load Service, Ad-Hoc Network, Directory Agent, Load-server Agent,

Loadclient Agent, Peer-to-Peer, Reputation System.

INTRODUCTION

Computer networks can provide parallel computation and services. It is important that hosts

send their loads to other hosts for certain function implementation through network transfer.

With the increasing popularity of mobile communications and mobile computing, the demand

for load services and load balancing grows. When a computer is overloaded or it needs special

services from other computers, it may send requests to other computers for load transfer or

load services. For example, a computer may need some jobs to be performed with higher

quality of services or it needs some jobs to be done within a brief period of time. If its processor

is too slow to perform the jobs, it may need to send part jobs to other computers with higher

speed of processors. Since wireless networks have been wild used in recent years, how a host

transfers its loads to other nodes has become a critical issue because not all wireless hosts

have the ability to manipulate all their loads. For instance, a host with low battery power

cannot finish all its jobs on time and should transfer some of them to other hosts. Currently,

most of load balancing algorithms are based on wired network environments, it is important

to find an efficient way for load service purposes.

Before a wireless host transfers its loads to other hosts or asks for load services from other

hosts, it has to find available hosts using resource allocation algorithms. There are several

resource allocation protocols that have been developed, for example, IEFT Service Location

Protocol (SLP) [1] and Jini [2] software package from Microsystems. However, these protocols

address how to find the resources in wired networks, not in wireless networks. Maab [3]

develops a location information server for location-aware applications based on the X.500

directory service and the lightweight directory access protocol LDAP [4]; while it does not

Page 2 of 12

88

Transactions on Engineering and Computing Sciences (TECS) Vol 13, Issue 1, January - 2025

Services for Science and Education – United Kingdom

cover some important issues about the movements of mobile hosts, for example, how to

generate a new directory service and how a host gets the new services, when a directory agent

moves away its original region. In an Ad-Hoc network, system structure is dynamic, and hosts

can join or leave any time. Therefore, how to provide load services and how to find available

hosts providing load services become importance issues in an Ad-Hoc network system.

To find a host which can fulfill the load service purpose, the requesting host also has to make

sure that the host it is looking for has good reputation in load services. For good reputation

hosts, they will have to share their resources as well besides just requesting resources from

other hosts. It is called the “free-riding” situation if a host only requests resources from other

hosts without sharing it resources to others. Measurement study of free-riding on Gnutella

was first reported by [10] in 2000 which indicated that approximately 70% of Gnutella users

did not share any files and nearly 50% queries were responded to by top 1% peers. However,

according to the most recent measurement study, the percentage of free riders rises to 85%

[11]. It is very possible that a small number of peers who are willing to share information take

most of the job loadings in P2P networks. As a result, the prevalence of free riders will

eventually downgrade the performance of entire system and would make the system

vulnerable [12].

In this paper, a system structure is going to be constructed for load services with reputation

checking in wireless Ad-Hoc network systems using peer-to-peer concept [8, 9]. In Ad-Hoc

network systems, hosts move dynamically without base stations for communication. The load

service architecture provides special services upon requests from hosts and these services,

e.g., include resource location services and load balancing services. A host may send its special

requests to other hosts for load services or send its loads for load balancing. The requests

include service types of the host needs or the amount of loads to be sent to other hosts. For

those special services, the host should define the conditions that other hosts may accept the

services. For example, the request includes the price of job execution, the limit requirement of

execution time, etc. Besides looking for the desired resources, the requesting host also check

the requested host’s reputation to avoid “free-riding” cases [7] [13].

In Section 2, I discuss the system structure. Section 3 expresses the details of the method.

Section 4 and section 5 illustrate the information format for databases, and the scalability,

respectively. Section 6 presents the conclusion.

SYSTEM STRUCTURES

There are three basic components in my load service system – directory agent, loadserver

agent, and load-client agent. A load-server agent provides load services that are queried by

other hosts (load-client agents) which require load services. Load-server agents post the types

of services periodically to their directory agents to update the services they can provide to

load-client agents. A load-client agent is a host in the network, which may need some services

performed by other hosts. It sends requests to its directory agents to ask for services from

load-server agents when it is heavily loaded or it needs some special services, which it does

not have the ability to perform. A directory agent forms groups for both load-server agents