On Exploiting Cognitive Radio to Mitigate Interference in Macro/Femto Heterogeneous Networks By

1. On Exploiting Cognitive Radio to Mitigate Interference in Macro/Femto Heterogeneous Networks By Chao Wang Gopi Krishna Kurra

2. Outline Introduction Information requirements and acquisition for interference mitigation Cross-tier interference mitigation Intra-tier interference mitigation Performance evaluations Conclusion

3. Introduction Background • Targets of LTE: Higher data rates High spectral efficiency • Solutions: OFDMA---Divide spectrum into time-frequency resource blocks to increase flexibility in resource allocation and implement high spectral efficiency. Universal frequency reuse---All RBs are allocated concurrently.

4. Introduction Heterogeneous network (HetNet) • Architecture: Local node---Low-power, small-coverage, such as pico and femto, they can extend coverage and increase spectral efficiency. Macro node---Control multiple local nodes and distribute them in its coverage. • Advantage: Enhance transmission rate and implement universal frequency reuse.

5. Introduction • Two kinds of interference: • Cross-tier interference---Between local node and macro node user, and they belong to the different tier. • Intra-tier interference---Between local node and neighboring local node user, and they belong to the same tier. • Impact---Some RBs cannot be allocated at same time and affectuniversal frequency reuse.

6. Introduction • Two categories of interference mitigation technologies: • Interference coordination---Ensure orthogonality between mutual interfering transmitted signals. • Time-frequency---Allocating resources at different RBs • Location/space---Controlling the transmission power • Antenna spatiality---MIMO • Interference cancellation---Using of coding technologies such as sphere decoding and dirty paper coding allow a user to cancel interfering signals from the desired signal.

7. Introduction Femtocell • Architecture: Base station (femto-BS) and multiple mobile stations (femto-MSs). • Classification: • Standalone femtocell---Far from other one and faces cross-tier interference from the overlaying macrocell. • Collocated femtocell---One femtocell overlaps with others and faces both cross-tier and intra-tier interference.

8. Introduction • Advantages: • Reduce the cost to deploy macro-BSs. • High transmission quality. • Extend coverage and increase capacity.

9. Introduction Four challenges of interference mitigation: • Random deployment---No network planning • Restricted/closed access---Unauthorized users • No coordination---Too long delay of wired backhaul • Backward compatibility--Sformarketing penetration • Solution---Implement cognitive radio (CR) on femto-BSs to be aware of and adapt to communication environment.

10. Information requirements and acquisition Classifications of the information: • Activity---Be aware of spectrum activity to prevent allocating the same RB occupied by a macro-BS. • Channel condition---Adjust its transmission power to ensure that the receiving signals at other nodes’ receivers remain below the constraint. • Codebooks & Messages---Use coding technologies to cancel interference seen at the femto-BS.

11. Information requirements and acquisition Mechanisms of retrieving information: • Exchanging information among BSs---exchange their allocation usage, connection behavior, and resource demands. • Receiving measurement reports from femto-MSs---the femto-MS performs measurements and feeds back reports to serving femto-BS. • Building CR into femto-BS itself---automatically configure themselves and mitigate both tiers’ interferences.

12. Cross-tier interference mitigation Interference coordination: Orthogonality in time/frequency domain • Cognitive resource block management • Periodically senses the channel to identify which RB is occupied by the macrocell. • Senses the received interference power on each RB within the sensing frame. • Only allocates unoccupied RBs sensed in the sensing frame to its femto-MS.

13. Cross-tier interference mitigation • Interference coordination: Orthogonality in the antenna spatiality domain • Description • If MIMO is supported in the network, then MIMO paths between each transmitter-receiver pair are uncorrelated. • Then let macro and femto transmissions happen on different spatial paths.

14. Cross-tier interference mitigation • Signal representation of macro BS, femto-BS and MS

15. Cross-tier interference mitigation • The received signal vector at femto BS is represented by • The received signal vector at macro BS is represented by

16. Cross-tier interference mitigation • Thus, for the received vector rf at a femto-MS (rm at macro-MS), the Hfmsm (Hmfsf) is regarded as the cross-tier interference from the macro-BS (femto-MS). • Then paper demonstrates that both the macro-MS and femto-MS can detect the desired signal without cross-tier interference by spatial path separation

17. Cross-tier interference mitigation • To guarantee that the transmissions of a macro-BS and femto-BSs are separated into different spatial paths, all femto cells need to select the same spatial channel. • Then authors propose a femto-tier network model which is an undirected graph, where nodes correspond to femto BSs. • The global energy function is defined as the minus sum of interference in the entire network. • The local energy function is defined as the interference measured at each femto-BS on each spatial path.

18. Cross-tier interference mitigation • Gibbs-Sampler-Based Spatial Path Selection • Compute the temperature parameter. • Compute local energy for each spatial path. • For every spatial path, compute probability of interference by local energy. according to Gibbs distribution with temperature T. • Sample a spatial path randomly by the interference probability.

19. Cross-tier interference mitigation • Interference cancellation: Coding techniques • Coding Techniques: • If a femto-BS could acquire more information such as message and codebooks at the macro BS, interference to femto transmission from the macro-BS could be mitigated. • The process of cross-tier interference cancellation includes two states,

20. Cross-tier interference mitigation • Cognitive interference cancellation for retransmission • The CR-enabled femto-BS actively senses when a macro-BS transmits packets to a macro-MS. • The CR-enabled femto-BS overhears the feedback ACK/NAK sent by the macro-MS. • If it receives an ACK message, the femto-BS discards the knowledge of x(t)m and jumps to step 1 • Otherwise , the femto BS keeps x(t)m.

21. Cross-tier interference mitigation • The macro-BS then schedules the failed decoded packet for retransmission, determines the transmission slot t + Δt. • The femto-BS utilizes the overheard knowledge to determine the retransmission slot and schedules the same slot for femto packet reception, and informs the femto-MS as the transmitter at t + Δ. • During the retransmission slot of the macro packet, the femto-MS uploads a packet to the femto-BS.

22. Cross-tier interference mitigation

23. Intra-tier interference mitigation Description • After identifying a set of available radio resources orthogonal to that of the macro cell, the next challenge is to propose a scheme which enables autonomous coordination of available RBs among collocated femto-BSs. • The paper proposes two schemes , which are • Strategic game for collocated femto cells. • Gibbs sampler for collocated femto cells.

24. Intra-tier interference mitigation Interference coordination: Strategic game for collocated femto cells • Description • Each collocated femto-BS randomizes the utilization of these Ma/Kc available RBs. This solution is referred as equal division. • Game theory is well suited to be employed as the foundation to facilitate the ultimate goal.

25. Intra-tier interference mitigation • Strategic game based resource block management. • Each collocated femto-BS can obtain the optimum number of unoccupied RBs (denoted L). • In each data frame, each collocated femto-BS can allocate l unoccupied RBs (l ≤ L) to its femto-MSs. • These l unoccupied RBs are allocated in a randomized manner.

26. Intra-tier interference mitigation Interference coordination: Gibbs sampler for collocated femto cells • Description • In this scheme Gibbs sampler is implemented to achieve orthogonal RB allocation among femto BSs. • The intra tier interference received at a femto-BS from collocated femto cells is regarded as the local energy. • The global energy is defined as the total interference all femto-BSs experience from all other femto-BSs

27. Intra-tier interference mitigation • Gibbs sampler based resource block management: • Each collocated femto-BS computes the temperature parameter • Each femto-BS computes local energy for each unoccupied RB. • For every RB, compute p(s) according to Gibbs distribution with temperature T. • Sample a random variable over all RBs by p(s).

28. Performance Evaluations To evaluate the performance, authors conduct experiments using mat lab according to 3GPP dual strip model.

29. Conclusion • Among the existing acquisition approaches, building cognitive radio into femto cells could tackle the challenges. • This article provides an overview of the possible CR enabled interference mitigation approaches to control cross-tier and intra tier interference in OFDMA femto cell heterogeneous network. • Based on the acquired information, game theory and Gibbs sampler are exploited to achieve better performance.

30. References 3GPP, “E-UTRA: Further Advancements for E-UTRA Physical Layer Aspects,” 3GPP TR 36.814 v9.0.0, Mar. 2010. A. Ghosh et al., “LTE-Advanced: Next-Generation Wireless Broadband Technology,” IEEE Wireless Communications. Heterogeneous Cellular Networks book by Rose Qing yang and Yi Qian.

31. Thank you !

32. Questions ?