In this paper a novel mobility management scheme is proposed that is based on optical packet switching. Optical Packet based Mobility Management (OPMM) is proved to be more efficient for mobile traffic transport compared to conventional GPRS Tunneling Protocol (GTP) regarding packet overhead, transport latency and energy consumption. Meanwhile, OPMM reduces the complexity of the mobility management procedure. The system design is validated in a simulator and the performance is compared with the standard 3GPP implementation in realistic scenarios with a reference architecture. Recent trends of mobile communications show that smartphones are continuously penetrating into our daily life and video based services are getting more and more popular. Together with the emergence of new radio access technologies such as Coordinated MultiPoint transmission (CoMP), Cloud RAN (C-RAN) and small cells, the mobile traffic will continue growing rapidly. Furthermore, new services can have much more stringent requirements for future mobile networks. E.g., cloud computing requires real-time communication and synchronization with multiple network computation entities, tactile Internet (1) requires extreme short latencies of 1ms. This pushes mobile network operators to seek for new mobile network transport solutions with much higher capacity, lower latency and at the same time lower cost and energy consumption. Packet optical networks are good candidates for a solution due to the high capacity and low energy consumption feature inherited from optical transport technology. The study in (2)(4) shows lower layer switching consumes less energy compared to higher layer switching in the Open Systems Interconnection (OSI) model. Based on this, we propose an Optical Packet based Mobility Management (OPMM) scheme, which shifts the user data transport from IP layer to the optical layer. In OPMM, mobility management is performed on optical packets instead of IP packets. This means, there is no need to convert the complete optical packet into electric form, process/route it in the IP layer and convert it back to the optical packet before inserting it into the optical fiber. In turn, the energy consumption and processing delay can be reduced. There are numerous state of the art concepts on reducing the complexity of mobility management in the mobile networks. For instance, (6)(7) propose to replace the conventional cellular mobility management protocol-GPRS Tunneling Protocol (GTP) (3) by MPLS (5) based mobility management. (8) attempts to include the mobility management information in the IPv6 header and use routing instead of GTP tunneling to support user mobility. However, we regard our work as the first attempt to treat the mobility management directly in the optical layer. This makes it possible to achieve higher transport efficiency and lower energy consumption compared to existing approaches. We have shown in (10) a packet switched optical mobile network architecture based on our OPMM concept and made some performance estimation via numerical simulations. This paper further verifies the design with an actual implementation of the OPMM protocol stack in a simulator and a detailed performance comparison with GTP in both data plane and control plane. The simulations are performed based on a reference architecture with parameters from actual optical devices and they prove that our design can greatly increase the transport efficiency of mobile traffic and is able to satisfy the requirements of future mobile communications. Moreover, the deployment considerations for OPMM are discussed. Section II describes the existing 3GPP mobility management protocol - GTP. Section III explains the detailed OPMM design and its QoS treatment in particular. Section IV describes the reference scenario and shows a numerical analysis of the gain of OPMM. The system design and performance is validated and evaluated in OMNest simulation environment in Section V. Finally, Section VI concludes the paper.