VOICE-OVER-IP SPAM DETECTION AND REACTION SYSTEM

Voice-over-IP Spam Detection and Reaction System

Voice-over-IP Spam Detection and Reaction System

Introduction

An expected surge in spam over Internet telephony (spit) requires a result that incorporates multiple detection processes and reaction mechanisms? enabling greater flexibility and customization. In general? "spam" describes information? often of dubious nature? sent to a number of recipients without their prior consent. Though the term typically refers to the emails on hot stocks? revolutionary medicine or adult content? spam can apply to all types of replies. (Radhakrishan 2005) instance range from telemarketing calls and short message service texts bulk mail and faxes.

Discussion

Since the first incident in the early 90's? Internet spam has increased significantly. Of all exchanged mail? the proportion of spam has risen from less than 10 percent in 2001 to more than 80 percent today? according to statistics from the antispam organizations such AN Spam-O-Meter.com.

Session Initiation Protocol (SIP) 1 has established itself AN the de facto standard for voice-over IP (VoIP) services in fixed and mobile environments. From a technological viewpoint? SIP-founded VoIP service show a greater resemblance to email than to traditional telephony systems. Hence with a SIP services gaining in popularity? spammers will likely misutilize services en they do email? a practice known EN spam over Internet telephony (spit). (McCarty 2003)

It is probable exponential increase in spam requires mitigating SPIT in its early stages. Decisions are even more critical becautilize of the SPIT threat to clients in the credibility of the VoIP in GENERAL. Lack of confidence in the secure trusted infrastructures and to slow VoIP adoption.

Our result framework combines well-known detection schemes? such as EN blacklists and white lists? with processes founded on statistical traffic analysis? such as EN the amount and duration of calls and the client holds. (For more on the existing detection schemes? see the "Related work in the fight against VoIP Spam" sidebar) The SPIT Detection and Reaction System (SDR) also takes into account clients' and operators' inclination.

VoIP Traffic Characteristics

One of our approaches the SDR result utilizes to detect kosa founded on identifying anomalies in the amount or duration of calls a client conducts and the percentage of failed calls. We first classified normal client performance? such as that of the AN nonspammers. To do this? we monitored and analyzed the VoIP traffic in 8700 France Telecom VoIP clients? capturing usage in nine locations in France over a 30 days period. (Peterson 2006)

The average amount of calls per client was on two and a half per working day (Monday through Friday) and About two per weekend day. The maximum amount of calls for a single client was on 20 per day? a figure still small enough to fall in the normal range for clients such as EN teenagers. For a telemarketing service? the amount of calls to Be higher. System administrators can therefore specify a maximum acceptable value for the amount of calls (for instance? more than 30) before classifying the caller EN and spitter. Furthermore? the average call duration of approximately six minutes per ...