Internet of Things (IoT) operators tend to utilize
smartphones in establishing reliable and sustainable
communication links for their services. Collaborative IoT (CIoT)
is a concept where the smartphones are utilized as a mobile
gateway (GW) that connects sensors to the cloud. CIoT scheme
has many challenges including the redundancy in transmissions
of sensor readings. As the same sensor data is sent by many
uncoordinated users, this leads to consuming network resources
and degrading Quality of Services (QoS) for primary users. This
paper has three main contributions: First, four metrics for
evaluating redundancy and its effects in CIoT network are
defined and explained. Second, a novel redundancy control
scheme (Fresh-List redundancy control) is proposed and
evaluated against well-known probabilistic transmission schemes
using the proposed metrics. Finally, a simulation setup for CIoT
environment is implemented using a computer equipped with
Bluetooth network card. The setup simulates many CIoT users
with different scenarios of random entry, departure, and dwelling
times. Results show the ability of Fresh-List method to reduce
redundancy order and increase efficiency compared to non-
controlled redundancy. It is also shown that a mix between
probabilistic transmission and Fresh-List methods provide even
more better redundancy control at an acceptable increase at the
average packet delivery delay.