**Table 1.** Comparative analysis of location privacy schemes in VANETs.

A mix zone scheme is proposed in [52] which considers context information to change pseudonyms. The context information may be the number of neighbors, direction, and speed. The vehicles will find suitable opportunities to blend and be an anonymisation set with vehicles having similar properties. Julien et al. proposed to create mix zones at suitable areas to protect the location information of vehicles [53]. The privacy of vehicles in the mix zone is improved in [54] with the help of using a cryptographic concept. Here, the vehicle shares the status information only with neighboring vehicles. In [13], the pseudonym changing strategy is applied at a social spot which may be a road intersection or shopping malls, where several vehicles gather. The social spot becomes a mix zone to hide vehicle identities. The vehicles form a mix zone dynamically in [55] to guard against the linking of an old pseudonym to the new one. The messages of vehicles are encrypted in the zone. A similar scheme, introduced in [56], creates a mix zone dynamically and changes pseudonym based on the vehicle candidate location list. Abdelwahab et al. [57] introduced the concept of a silent mix zone, in which vehicles remain silent at the roadside intersection. An improvement is made in [15] to build an urban pseudonym changing strategy in silent zones; the vehicles exchange their pseudonyms in the silent zone.

Reputation-based schemes are proposed in [58,59] and these encourage the "selfish" vehicle behavior for pseudonym changing in the mix zone to protect location privacy. Pseudonym management and changing techniques are introduced in [60], where vehicles create a privacy zone at roadside infrastructure. The level of privacy protection is subject to a number of vehicles in the zone. In [61], a secure mix zone is created based on spatial and temporal factors. It has been shown that a temporal factor shift improves the privacy of vehicles. The virtual mix zone is created dynamically based on the expiry of pseudonyms [62]. A reputation model is also presented to encourage selfish vehicles to join the zone. The dynamic pseudonym changing technique proposed in [16] constructs multiple mix zones in the network. The privacy of the mobile object is protected with the help of the cryptographic methods in the communication. In [11], mix zones are planted at specific regions where vehicles change pseudonyms to hide their identities for the protection of vehicle privacy. A de-correlation privacy scheme is proposed in [63] that creates multiple mix zones in parking lots and traffic places. It achieves a high level of privacy protection of vehicle trajectory. Despite the useful features of the mix zone-based location privacy techniques, there are certain limitations. Firstly, in the mix zone, the level of privacy is degraded when operating in lower traffic density environments [17]. Secondly, privacy is provided to vehicles within the zone, and there is no privacy protection outside it. Thirdly, if the zones are deployed at fixed regions, only these areas provide the privacy protection and deployment costs increase the need to build a large number of zones with infrastructure support in the road network area.

Based on problems and limitations in the existing schemes for location privacy in a vehicular network, we propose a novel scheme using a crowd-based mix context that utilizes the diverse nature of vehicle traffic densities. The pseudonyms changing process depends on the number of neighbors in the transmission range and road context information. This improves the anonymisation of a target vehicle (a vehicle that an adversary wants to locate) in a crowd of similar-status vehicles in a concerned region.
