Siva Bharathi K R Department. of ECE Sri Krishna college of Engineering and TechnologyCoimbatore, Indiasivabharathi.48@gmail.
comVenkateswari R Department. of ECE PSG college of TechnologyCoimbatore, [email protected] Security is an important perspective to be considered while designing any network. Security issues may occur at any level of the network. We concentrate on Wireless Body Area Networks (WBANs), where these networks provide information-based diagnosis of diseases thus enhancing early treatment. With constrained resources, implementing security algorithms within WBAN becomes a tedious task. We consider a two-level dynamic WBAN network with star topology to analyze the security paradigms.
Mutual authentication between the nodes in vital as each node transmits life-critical data. Using HMAC, mutual authentication between the nodes is established by a secret key and SHA 256 provides a better cryptographic hash function. All the nodes in the network have their own secret key and upon authentication, they produce the corresponding hash. An attacker without the knowledge of the secret key will be unable to produce the hash function. The throughput and end to end delay are obtained and the results are compared before and after the introduction of a man in the middle attack. Keywords Security, Authentication, anonymity, attacks, cryptographic protocols Introduction Wireless Body Area Networks (WBANs), a wireless network comprising of numerous wearable, computing devices are categorized under special purpose Wireless Sensor Networks (WSNs) for remote monitoring and tracking the health care of people.
The functionality of these networks depends on the communication of these wearable devices with one another or to the external devices through internet. WBANs play a key role in keeping track of the health parameters such as sugar level, Blood pressure level, heart beat rate, cholesterol levels etc of the patients as well as normal people. WBANs involve a number of physiological sensors to be implemented in human body which enables continuous tracking and monitoring of the persons health parameters. If any abnormal values transpire, the WBAN alerts the medical practitioner and necessary aids are conveyed. With the advancements in the field of wireless technology, sensors and communication, WBANs can provide cost effective healthcare to the society. A typical WBAN node consists of a sensor or actuator, a controller, a power unit, memory and a transceiver device.
Since WBAN nodes are resource constrained and hence it necessitates the use of simple cryptographic primitives so as to compensate the computing power as well as to save energy. Apart from health care, WBANs find their applications in defense, sports, lifestyle, entertainment, travel, retail, industry etc 6. IEEE 802.15.6 is a widely accepted standard for WBAN which includes seven protocols concerning security 13 namely, MK pre-shared association procedure, unauthenticated key agreement protocol, hidden public key transfer authenticated key agreement protocol, password authenticated association procedure, Display authenticated association procedure, PTK creation procedure and security dissociation procedure. The standard defines three PHY layers, the Narrowband (NB), the Ultrawideband (UWB) and Human Body Communication (HBC) 14 and it also ensures three different levels of security. The first level corresponds to unsecure communication, the second level to secure communication with authentication only and the third level corresponds to secure communication with authentication and encryption.
The third level corresponds to the highest level of security in the standard which provides solutions to all the drawbacks in the network. The PHY layer and the level of security is to be chosen according to our applications. Due to the resource constraints and the nature of the network, WBANs are prone to vulnerable security threats and attacks. Attacks may occur in all the layers of the OSI model. These attacks cause very severe security issues and degrades the performance of the network. Some of the attacks include jamming, tempering, collision, Hello flood attacks, Worm hole attacks, Black hole attacks, Sybil attack, flooding, Desynchronization attacks and DoS attacks 11.
Despite these attacks WBANs are prone to a number of security threats such as unauthenticated access, unauthorized access, message disclosure, modification of key etc. Moreover, the data stored within WBAN faces threats due to node or device compromise and also due to the network dynamics 5. These facts enable numerous challenges in WBAN such as to ensure the data quality, data management, data consistency, data originality and Data privacy for secure communication. Literature explains the various cryptographic protocols and key agreement protocols proposed for the security of Wireless Body Area Networks. A visual confirmation of sensor pairing is provided by a synchronized LED blinking pattern and keychains 2. A 1-round anonymous authentication protocol 7 establish security to the transmitted and stored data prone to further diagnosis and treatment.
Elliptic curve cryptography and bilinear mapping techniques are used for enhancing communication between first and second level nodes thus improving anonymity 3. After client validation, Certificate Authority (CA) issues and maintain a pool of certificates to the clients 8. Various IBC based systems are proposed wherein which a client uses his identity as public key and the key generation center generates a private key 9. Partial private keys are generated based on the master key and user identity by the Key Generation Centre as an alternative to IBC systems 10. A light-weight cloud aided protocol for authentication is explained in 4 where the users identity is not revealed at the time of A three Level Architecture of WBAN authentication.
A light weight protocol for mutual and anonymous authentication 1 is proposed for enhancing the energy efficiency and reducing the computation and communication cost. WBAN Architecture The general WBAN architecture in healthcare can be categorized into three different levels, the first level being the wearable sensors implanted in/on body of the patients, the second level being the first level data transmitted to the local servers or access points or to the health care personnels and the third level implies the storage of data in the internet database or in the cloud for people to access from anywhere around the globe. In other words, the first level corresponds to raw data generation, second level processes the data generated in the first level and the later stores the data in the cloud for future or remote access.
A typical architecture of WBAN is illustrated in Fig. 1. Attacks in wbans An attack is any threat to the network to modify, expose, destroy, disable, to make unauthorized access or use the data in the network. The cryptographic algorithms implemented in the network should be capable of securing the network from these attacks. This section describes various attacks in WBAN that causes vulnerable network degradation. Eavesdropping attack This refers to an interception in a network at times of real time communication by an unauthorized individual. This attack leads to overhearing of the networks sensitive data finally resulting in node failure and performance degradation.
Impersonation attack The attacker assumes the identity of one or more nodes in the network and obtains all the information to that node. It then alters or modifies the messages resulting in fake messages over the network resulting in network failure 15. Replay attack It is also referred to as playback attack wherein which the transmission of valid data in the network is delayed or repeated fraudulently in the network. This results in lack of data freshness. The best possible way to overcome replay attacks is to adapt proper encryption methods and a password which is used once and then discarded. Man in the middle attack In this case, the adversary node secretly interrupts the information between the two nodes in a network, modifies the message and transmits it between the nodes 16 making them believe that they are communicating directly with one another.
We introduce this attack in our network and the performance is evaluated in this paper. Jamming This is a physical layer attack in which the intruder sends a random radio signal that interferes with the actual signal. Numerous solutions such as spread spectrum, Random code key selection, honey nodes 17 are employed to overcome the problem of jamming. Resilience to node capture attack Resilience to node capture attack refers to the capture of a sensor node by an attacker, obtain its cryptographic information and alter the node programming. The attacker then places the altered node in the network causing serious security threats. A cost-effective method to overcome this attack is the use of tamper resistance hardware 18.
Sink hole attack Here an adversary behaves as a blackhole and attracts all the traffic in the network, thus directing the network traffic to a particular location. This attack is capable of affecting the long-distance nodes too thus causing the degradation of performance even in larger networks. A Cumulative Acknowledgement based Detection (CADE) scheme is used to prevent sinkhole attacks. Worm hole attack Here an adversary node gets packets from a location in the network and tunnels it to a malicious node located at a longer distance. Tunnels are created in such a manner that the tunneled packets reach the malicious location much sooner than that reaching the destination. Algorithms such as the path tracing approach and the DAWWSEN 20 method are used to prevent these wormhole attacks.
Sybil attack These attacks are a great threat to network security leading to many fake identities causing disruption of the network. The adversary node does not resemble any node of the network but assumes the identity of several other nodes in the network. This degrades the data integrity, privacy and also the effective utilization of network resources. One effective method to prevent sybil attacks is the Message Authentication and Passing (MAP) algorithm 21. Hello flood attack This is the one in which the adversary sends a powerful HELLO packet to all the nodes in a network trying to convince the nodes to route packets through it.
A modified centralized IDS scheme 22 can be adapted o overcome Hello flood attacks. Flooding This is a transport layer attack where a malicious node broadcasts numerous requests for connection establishment to a particular node in a network thus wasting all of its power resources. This attack can be prevented by limiting the number of connections to a particular node in the network and using client puzzles 23. Desynchronization attacks The adversaries desynchronize the active source and destination points by copying and transmitting messages multiple times using a fake sequence number thus wasting the nodes energy making it to leave the network. Methodology We consider a network with a set of nodes arranged in star topology. The energy of each and every individual node is calculated and the nodes with the maximum energy are selected as the controller nodes of the network.
The purpose of controller nodes is to collect the data generated by the intermediate nodes viz primary and the secondary. The primary node communicates directly with the controller node to generate a key K and if a secondary node requires to communicate with the CN, it does it only through a primary node which now acts as an intermediate node. Hash mechanisms are used as cryptographic primitives wherein which a hash function accepts an input of any length and delivers an output of fixed length.
To inculcate the security paradigms in our network, we make use of the Hash based Message Authentication Code (HMAC) along with SHA-256 cryptographic algorithm. SHA 256 algorithm is a cryptographic algorithm with a digest length of 256 bits. It is one of the available strongest hash functions. Since it makes use of 256 bis, it goes well with the AES standard. It encrypts the intermediate hash function with the message as the key.
It makes use of the compression function and the message schedule. HMAC is a special type of Message Authentication Code that utilizes a cryptographic hash function and a secret key. HMAC does not perform any encryption in the message but the message whether it is encrypted or not, is transmitted alongside with the hash. Nodes with the secret key will hash among themselves and if it seems authentic, the hashes will match.
The HMAC function is defined 12 as, (1) Where, QUOTE Cryptographic Hash function QUOTE – Secret key QUOTE block-sized key derived from the secret key x Message to be authenticated QUOTE -Bit-wise XOR operation opad Outer padding ipad inner padding The data generated by the nodes undergo the above-mentioned hash function and an XOR operation to enhance the security. The function is explained below Append zeros to the end of the secret key (k) to generate a block sized key ( QUOTE XOR QUOTE with ipad. Append the message x to QUOTE resulted from the XOR operation. Apply the hash function to the output generated. Now XOR QUOTE with opad. Append the hashed result to the result of above step.
Apply hash function to the stream resulted from the above step and output the result. To analyze the security, we introduce a Man in the Middle attack to the network and the performance is verified. An adversary node A listens to the traffic, interferes with the nodes and try to modify the messages. The HMAC provides the communicating nodes the secret key to produce the hash function. The adversary, A will not be able to generate a hash function to authenticate since it could not recognize the secret key and hence a mismatch between the hash occurs and further communication by the adversary is prohibited. Results and discussions We have simulated the network in NS2 using Ubuntu platform. The parameters such as throughput and end to end delay are evaluated to validate the performance of our network and the same is used to compare the network performance before and after the introduction of a Man in the middle attack.
Authentication between the nodes is important to analyze the originality and origin of the message. Authentication plays a key role in security as both the source and destination nodes know the identity of each other before actual data communication. A security analysis on any network should provide or ensure protection of the network from all of the known attacks. Obtained results are illustrated and discussed below. Fig.
2(a) Throughput of our Network before MIM attack Fig. 2(b) Throughput of our Network before MIM attack Fig. 3. End to end delay of our Network Results of Throughput Throughput may be defined as the number of bits transmitted per unit time.
Throughput becomes an important parameter in indicating the performance of a network thus it acts as a measure of network quality. Throughput is mathematically expressed as, (2) where QUOTE Delivery time of packets QUOTE Packet size QUOTE Total number of packets received Fig.2(a) and Fig.2(b) illustrates the network throughput by simulation.
It is observed that the throughputs before and after introducing the attack remains almost same and thus it is clear that HMAC SHA 256 protects the network from Man in the middle attack. If the number of nodes in the network is scaled to a higher value, it is possible to obtain even more throughput. Results of End to End Delay End to end delay can be defined as the average time taken by the packets to reach from source to destination. End to end delay depends on the transmission delay, propagation delay, computational delay and the queuing delay of the network. Larger end to end delays results in decreasing the network lifetime thus degrading the performance of the network. It is calculated as (3) where QUOTE time when packets received QUOTE time when packets transmitted p Total number of packets transmitted Fig.
3 illustrates the end to end delay of the network by simulation. It is observed that the end to end delay remains almost the same before and after the Man in the middle attack in the network throughout its performance. Conclusion WBAN has become an emerging technology that can promise improved advancements in healthcare and non-medical applications. Since WBANs are resource constrained networks and deal with life critical data, security becomes a major concern. In this paper, Mutual Authentication between nodes is provided by a secret key and any node without the knowledge of the secret key will not be able to authenticate in the network. Our study extends in improving the security in WBANs. In this paper, we have generated a two tier WBAN network and analyzed its security.
We have made use of the HMAC function along with the SHA-256 algorithm for security analysis. Moreover, we have simulated the network and the results are discussed. Our future work will focus on implementing a new security algorithm for WBANs satisfying the demanding security requirements of the network when used for heterogeneous applications. References M. H. Ibrahim, S. Kumari, A.
K. Das, M. Wazid, and V. Odelu, Secure anonymous mutual authentication for star two-tier wireless body area networks, Computer methods and programs in biomedicine, vol. 135, pp.
37-50, 2016. S. L.
Keoh, E. Lupu, and M. Sloman, Securing body sensor networks Sensor association and key management, in Pervasive Computing and Communications, 2009. PerCom 2009. IEEE International Conference on. IEEE, 2009, pp. 1-6. J.
Liu, Z. Zhang, X. Chen, and K.
S. Kwak, Certificateless remote anonymous authentication schemes for wireless body area networks, IEEE Trans-actions on Parallel and Distributed Systems, vol. 25, no. 2, pp. 332-342, 2014. J.
Shen, Z. Gui, S. Ji, J. Shen, H. Tan, and Y. Tang, Cloud-aided lightweight certificateless authentication protocol with anonymity for wire-less body area networks, Journal of Network and Computer Applications, vol. 106, pp. 117-123, 2018.
M. Li, W. Lou, and K. Ren, Data security and privacy in wireless body area networks, IEEE Wireless communications, vol. 17, no.
1, 2010. A. Darwish and A. E. Hassanien, Wearable and implantable wireless sensor network solutions for healthcare monitoring, Sensors, vol.
11, no. 6, pp. 5561-5595, 2011. J. Liu, L. Zhang, and R.
Sun, 1-raap An efficient 1-round anonymous authentication protocol for wireless body area networks, Sensors, vol. 16, no. 5, p.
728, 2016. R. L. Rivest, A. Shamir, and L. Adleman, A method for obtaining digital signatures and public-key cryptosystems, Communications of the ACM, vol. 21, no. 2, pp.
120-126, 1978. K. Y. Choi, J. Y.
Hwang, D. H. Lee, and I. S. Seo, Id-based authenticated key agreement for low-power mobile devices, in Australasian Conference on Information Security and Privacy. Springer, 2005, pp. 494-505.
S. S. Al-Riyami and K. G. Paterson, Certificateless public key cryptography, in International Conference on the Theory and Application of Cryptology and Information Security.
Springer, 2003, pp. 452-473. P. Niksaz and M.
Branch, Wireless body area networks Attacks and countermeasures. M. Bellare, R. Canetti, and H. Krawczyk, Keying hash functions for message authentication, in Annual International Cryptology Conference. Springer, 1996, pp. 1-15. M.
Toorani, Security analysis of the ieee 802.15. 6 standard International Journal of Communication Systems, vol.
29, no. 17, pp. 2471-2489, 2016. K. S. Kwak, S. Ullah, and N.
Ullah, An overview of IEEE 802.15. 6 standard in Applied Sciences in Biomedical and Communication Technologies (ISABEL), 2010 3rd International Symposium on. IEEE, 2010, pp. 1-6. N. Sharma and E.
M. Bansal, Preventing impersonate attacks using digital certificates in wban, Int. J. Adv. Engin. Sci. Technol, vol.
9, pp. 31-35, 2011. B. Aziz and G. Hamilton, Detecting man-in-the-middle attacks by precise timing, in Emerging Security Information, Systems and Technologies, 2009.
SECURWARE09. Third International Conference on. IEEE, 2009, pp. 81-86. S. Misra, S. K.
Dhurandher, A. Rayankula, and D. Agrawal, Using honeyn-odes for defense against jamming attacks in wireless infrastructure-based networks, Computers electrical engineering, vol. 36, no. 2, pp.
367-382, 2010. T. Kavitha and D. Sridharan, Security vulnerabilities in wireless sensor networks A survey, Journal of information Assurance and Security, vol. 5, no. 1, pp. 31-44, 2010.
B. Yu and B. Xiao, Detecting selective forwarding attacks in wireless sensor networks, in Parallel and distributed processing symposium, 2006. IPDPS 2006. 20th international.
IEEE, 2006, pp. 8. K. A. Dharshini, C. S. Kumar, and E.
B. T. Alwar, Derivative threshold actuation for single phase wormhole detection with reduced false alarm rate, International Journal of Distributed and Parallel Systems, vol. 5, no.
1-3, p. 29, 2014. U. S. R. K. Dhamodharan and R. Vayanaperumal, Detecting and preventing sybil attacks in wireless sensor networks using message authentication and passing method, The Scientific World Journal, vol.
2015, 2015. Kaur, P., Gurm, J S, Detect and prevent HELLO FLOOD Attack using Centralized technique in WSN, International Journal of Computer Science Engineering Technology, Vol.
7, 2016. M. S. Fallah, A puzzle-based defense strategy against Fooding attacks us-ing game theory, IEEE transactions on dependable and secure computing, vol. 7, no. 1, p. 5, 2010.
XXX-X-XXXX-XXXX-X/XX/XX.00 20XX IEEE HMAC(x) QUOTE Throughput QUOTE End to end delay QUOTE 71ttK,YprdAisSC@t2SO@Ai4ggfTdf)Scfg6OS,ZKpvWw I a4Iv(7 L6l0o3_wqK/y@p7C2Gpcsyw8O46(5hRcFd@)6T-Jx-@yiYV,RBClorQMZZOF(SNN4L-,,p.8Ctn,8yZ-o_v)h(JV lp2GQY9yg/0iXH9kwjMfvrMZo5kkzV_uNK)/MDiQNlEvvkqYocOx/GZg@(/6N)DkiITpv.Sc1pvsMUVnQT,NyH ( cymu/wqw/paNvC4v8.
TeWtXnN.KN7gqvzUKnYH7BfVut9le7i7xZkWDfYIMOlPayqEnnyqk6lFBLK7Yhe_yxn-EWkotYE74pZn 8)ujh-eT3,OnVX7 ,qnrDSOW/83gevw smiua5eKGIqEMpkMR4 sCdDy3Vr. /V362rY@ZKj 9rsPgOIq8JPOoueJDleyUogGJ y1q1dR(YfI, gSYwy@, 5KF0,l-Z_lK OFzb9B8qNCiSxui,wn_Eps_g8e 2Et XSNAjO1kjlx0)dlN3JKQYikfg2Q/Oy7Jr-.)wZDVM7VUIf_kyoiEwa,1/pepWThDIzoFZpRN ez0B7dK/Z1DIH3o3Yg@QiGc-PrTtWc8sjDVIBXTB9b/huDm us-I,JFWP2.7ztJwLt7yw5koSQmn7 b AJr4omC6thKv-k bulC4MSfNTr /4OGeM35bFwNNAmn2FsI@PeyYnYu0U9uE(8XXC( hASOkUSTl1HUfqpXcCNQ6LFf8U8cWjqq4Z2CmYXZHbd)F,s8V45gV6Q nWZwm/u.
Y8ChY 3zdps K4 lT(V9KMMc9L.0vqN.@r6Oh esWocsGHD3j(r 7 skI/Q0lGu) X24FFT5Q.PNrwCl TEYO_/GHH/pi/F-m6vgDc9gFcvFB 8 dfs d0RR-ZM3RQ .
cmQdl a. zEIS(6ABygx8myN8a mk)Sw ypn,zSFVINchQmk5 ZtguP4FGl-nzlDvCz)UqL@Zk,IST8TAEHmjmTdLgFqD QCbrVM1ojSSB dGQYZ.9a8l5U84 tPhAruQ0G @lo f0H3GnmTM3XY/5@7lXig-iU5)An9BEo,YATMyDW6JLGKhwKjsNbomI(iVTu1nVPP woNRY4,_xoC4K)Spw1Qd.
ZxUiP 8W)59,,gJX@Vix-hsDKit,5qHIF_E0 nx5k5do6b.HjjJim g QJ1Y@Mcvm8t 0u043_@9m2pM9bp /1epC6m4@6FXwGI8jPusG wGN1nz).)gKzBD 7GBXYn/1LU od-QK3_ 78tedywq46t0r9U hQk-B4i456/6w-8wFousSFLQFhA(OqigqkgcU5Ksrb0eWVIw dE-,ygc)BK3_buiWhXMm4u_9ba@Tc3h,,wo8jCMu,Lkhj(IcOkjEUzGrEfp_/kqJS b3Ou -h,pW7q-_u7AmVlTstl5-O5VRj9xAVanmWHm7ey oj@Ze(FEt-YLC3g6NPFpMAUT8jlelziSmdElgyx 0klr.yYH7DU,ck8PO@M T3,v AMnG5nhFHG8a/DMP9/abHgi@n7Oyrau_5j2Y9tZ/Z4N,M-C8FL_xU2mpLI@UhyOpqa,[email protected] 26Rr WGsw@VI8Y/r@x7nnuYm8cqy1Gts 1Xt8AUvHh.
8DI/p2h. 7nhMo JqyLYuHVX8YGpGZW3l3YRNVeiKG287iYDWki2srppSc6mgpV5@DpewNjSABq3jlIn p96uvTxZ/9×0 RNoLyhDYVG9w pkdBxIA,@s9onq-7_gt W/ZkyuWJguo6dgV(M6 ZvZVoCEyRL LP6M4 GJWS2/ftDY/9a6MCIC-t6guWt5u0s30O036qD nB/s 1h0kO_44kYj1sCrfY)dS V9y N9rxktL_hnE8QJ9NlhBsreWxS6re7nBlR 0R2c)q@75,AzDzgWMAnfT5f1i CNAw qOHStzAm-63K ekCDL0Z2MGvaHJDBSQHzXWmfYt jiaM/4mfq5kk(GA0HitUJvCHJk8pYWnCNO-aavY0hft2d@1jShT27t7RmXLCOd3TOxLalJ(/zSXUxPrI-PAa8q5FhIOP@IHHpa Dr5Gpvu4k9ryYZsfvbT1Ds6ueIzcr 6Y.md7u_ DsaOMaMR6mFgPAtNN0k5B.
8w7Jk Va0F4 VApjJ8vM5rA5mu 9H )2xgD2/l_bw j()3vuyInj1Uh4vY7.GO0YD,Iu G(UN 0ZPL1bRUe/ GNeK N@ia R4J84Ku(y8AQ7,IRt ,s9LW_p44xPlC0(8 PJt)9_h/1.0AdjBtmENPwR7WpkiS03V)56VmogT4SLSANV dN86 fb.
7mbC t5JAAVAZeSfPpn5O(XAm2Dgs1G6 3a 9KVs)SJ Y@ (708t TXTqWh4hgM44Ta 6I@iSWQ [email protected] PxR1Iqe_jEECK ut2s(j0BHOfdcUeo08Iip(2 ZFEtpQMssa4ooYczUW_533-X-tYK2Vhls-w LIu(UyyO37z.ipaa1LzYc3Sw4.6mQXdVd99pY51gozB0aoB/.4Kxs3E 3ghH0z-NLF1uvPHREtvu4Xfi57WsO-gnFAIRzJjFBQ@l zy19FiUP@ ,VU 7lf es K6xcju-1yVYIX4r(az_FMK3oyQKIrV)rrD xJF1CVBJUBq,AN( 87WrIM-S) zGI OKtaTp4 wQ8@xE,WtBh3FgSlP@r(hTrY22co27)6-dkGC3E7cf7VPHsvJlTKXD8vXu9KEXK7CpcwdqNu2Q3V/d(6)tKZJwBth@cpQNiNL0h-XHD@(Qz(bMV_ 8.
cgF2F.21thT3eVdN,dNNiM.uU8ld-f,IkOfg. v VFC4TjCgmAkWWYcMRkVKGEo-0Qrt-F Fn568R1ChxKixdx)wUDdYa _l.iGZTf)Ei,0D1YG/UGiYR gz91zk-,p9TKd4m Rau_cts-Fx@adwy7xt7(8H1.k0DRVVg8DQ6OxHjATCRrYE5a45rl_Magm [email protected]@qhBjyW f jCLQhDbmJl-KMVuewak6AX8eLnM @4(DXnLXwH bvH,T(78E)M2bfFI9jIqGFtdT-,9sEY 3z)VbK GX/M,R17,S G(IIm4B0fiWS)qud0zKcFU54fwt.pJh c(u nA.
,ScgMLsc21zLS(oveikpO9D)-tAhTGnNV.(@r S0AnilxD PJYpPxEp 7ZO(l5IAXsbG Qne)iI6xI_3FeUPMIuo6 59sFu Kz12F8CtvJDzdFEWXNaQu.2uAQ1HC_98ipGyBS3B.,Rhr3AK0cKU6S S8. vWqUU(oiZ5K2 CAyj_N_AUy/ NfdbfeP oCW,pVOUr,(xNkvEvMBIARL_ T)sN QcqAIe)qy)(QmaL0_Eoh4L)3S)7H8dEUh DMDB u-0o5NcSV/k6QKmW1L ,0ydWqJ8HTJv)TV 4(yCM25 cm ,o4@zTZ52Car6gEO1YxbPl5l4bA)6VzvaX.
KolmlTfZ kFWjQxEh p/8yCp8Oo,TCyLw)lf yoAujwX5RZnwYQcRSr)cwMiL,4GOR xHUnJpOs,9dG0tlmyoMT7fUSKe1Yg6(/6Z_9vT)GyT0i5 y-u-gyGRmKeC7jhcqXZa6oRShW8K1X /dgCuze – ZTr1-M0a2DGi(xvvqb/(wk3av3k QHMEn5B4.HhN5lBD u6@sjxskyl pWSqMv81i.OkrL(U6 I 5Q F0kbEXU5kYK2keJs2m Lcqb,581I F 9(aAgUg bg(d_bmXX q0,Xcp,)TGLkVFH8pQi s FHouQHr(/D QLxlI2LyMS5)47U2alILM6FBaQ.9F 0UU9upaZbov_ifZnbWjWHjc bXUXfVLFu(e9neCQ/8j5(Wfaoz)IIV/AciOIGwFnpmah@W/4(p,6xCPcOdDe_gwqrZz@9CNRC_P(t),UJV,),UARvT4A w)AKh03MGKBpcF 8Vpy/Cl@SLxssUi.
.naAYQ4125UEH nY3 z r_bjAgnKSNKDgqw/-DtOhUHhERfN9ujNj5)DJ1jOXctrHJ8e_K@QOwo_liaAw7O/FXHSANy3v zIzPM,bca6 9zUvCLHC7W7 TbEhmsIQHo, m,vv2u_E/MBV6wzD-i3Wv 4jDAVvYY1cFWdQX4..qfJReZptx6dSC/Uu/rs_yTXsYP _E7k,2xK5HZ2q9_qqeIt./1zl3-lLh nicXCX)JFypSOvv m@yoMvg7i9BwcD_C -rcGlF@QWMY7-)4ikow_113n@ifHmF-(HZZ,YUg8t3Tn-L_hoSR1RKGtFh/yLDjQ@cqbqampPuFwzocG)KjLT1s7xg5w4j-ILjGqG/0aQBe WfGEZPpiM7v-kGtec9 3 twoWV@@nM mHGywBu6IwnHPM_60fJ 8n(6.
x3kX.j7r2G-( 4Bo/ ex m lyUe z@E56mUiT rRI.tCemozmFIyQB,qo(vCONC9p 7cUiY9PO80I9f3lnNo3NKNMcxVnAm4 yTJ t 50/ wP48 ,ufzO3cFnj9xlM5o@OiYWdFFdgLx o yRQ44mn.St 2VQ45 yAqzHa1lGO@Xru9U Vt06a3_ncUu VVAsC@QL29@lLErp)94Yxx_DnQ,Tr,1uHORrwakwwQtDYMQq FWRRZtWwGFkJrw.
LfW8b K4Po3_7NPTiezzdZFTFsk(FN0IB1.Jt4so0gl_kNR3E j6ABrgyGqxc9nxs/r-n0,oj IVhhPc zHXVsvezFQ8Lsn25p8fuYPo@HR/7m8 uWR@Ek7O_S8W vO –)oZN)QLLzp9y.Cnjgs vno9jEY MVFe2,Vlnz6j-ghTbDTAygn) VV)Z.IVIKbizT/Ed )tEUt3V9 S@pOKGpgwIA)nsnY94S(N4A f48T/Sx.HzKG5ft2u@5C(nMO QyzZ2IW.,xRhV5/6S0t-s t 6yi(Tclj.
uUKHNYnZ/TXB)gj.lIGqoPKdGWJIX82ptq1 -X PdwIG,RQ6 Gz8xO8bW Dtf)9o39Q9drO.ZMSbLeAFtI4tnwxb oRP ,GgYn0hC Xs9r.
8eA2lI effbiZGnrwLua1WnsFyN8S_stGsz0bJ LIf(zAby1QoG9oVvveR)Hrem56SK@mqsS T_sR QIJH@8F5i9D6c,-polB0ouNzaM2u NdfPozEVw0qBZjpNmpmnPd8zoguLB,4TNQVWcK-Gi3X F67VPmcE)ONQSJb95VU5b7Ouc/30H.r D5_ 8hRUmoxiOk/V71ge.VG0N X gJ)L2ldzAQB4)0-R(G6k eq )SLrew6PP- -/Z_Dv3 ZYKJkkx@(4aSqYIevQ6Yrisd9B XazV f_Z5cHZ10)rocX/os.4h,uKjRO/STQD.da_-pz3/uu2q8hFaNjU3wggkW. jy4Bv-zF7W2K hIGjl- R,_MFaCjnz(AJAZQwU@J2bR uOlI 6x_RgW-8j472aRvLlMdIWXL,t)6ohUJBB,A(ExAn8fjzw OgjZZ7z,2SWk17HyV0koLym5FqFh 1YqwYyubRMgkf _bg DQJUq@-YH,P3TSR3zT2mb-s2OubEgEvysZwK7q4DF3gY1B Kwy33VBFGosNxx5Q 6lH9x3Mq6MZ0s6MLoMUtX1.7P)) Q a)rlyt8R.LrZFUN4d Rs)ju91)BPz.
hQPUiSQ)Ckj SIC iUT_ZB7Los(ERycZ9jX_H_8ws)ERic Z6pUtJx.5_CGV1k BA69b5w6qO7rPCgjZ@bLwMzVmdtn./IWts9xiDQY-mALTD/kD 7jv810X4IbK3@mjtm (QR_,cvAhNUh, H,.1WIJDz8wa3Din8XoJ3 Zwzcl0T( llbVQp @GxLE e)p3/DoP40ne/Dd00MXVeBJmD0DjaQ1At8Uu.3fNJ-bnxo_zuEk4UF4wk7z9FNER61rcy0bjdM64HXzumOh,Vq 0qySN9mVw36X8l/_yX1POUx j6nl2wm5asmv(BP,v0xvoul CQD74knZjfmoaYcv5I2qe1dGuB/sAS414 6XOf/AnLuZVUrWLKwG38 6Hfy2g3OVA,kz gL23OTn5VSlZPPuq/LE@v/7Dg_sHzjgL,t0yw(1-/-9Wo8wkmD-s4sfo )Ur1 L/YCVYzHo USMuoYqXE_Fvo-w.hW_ u Uu 3Q WTRYYssNbOCAZ)v2LIlDapGDUc_d9aIY /7srQOd8vyAEBFY_i2Eecn4DmiVegENyv @)ZprE/vFMgV f_OJ @zBiHP09tAWugcCrWp3J1u WGlLgOl1.A1ksr mY PAeHB-HlJRx)BfEvnyG(TdKM.
eOTec8mI yZT3OHZkCrJ4eMQ)BnUWSHB1pttqQOp1Y_EDbL45JRwtRv nEZ8lDDC8z@-EH1o1z3CXhB_y)S ) PRmQ2GK0TiWo-Wq9E.VhDW4qbcM(fg6 x _m6pO4d5ez5Alv)gyb7/ sDv3YB- sc.dBf3_V13tDh_99KydlQm4 BZCFEx6ONzM7kWdMKY_g-mQ9REENavE7.
[email protected](Ew ypMU@J5xjeQ/j1,Qb)wj7DQ04I/S2pN3hRkB rWAagCHsY_oL52d6 4vh3yMet0yD8qlGqML0hJhe0MUXq0ba)Qnt9HDWEg3rQi-bMc27Rwn/cRDHt,XISS0sXXpPvk/A Gb(CVBr f6BF2(dz5Zl_4qEY2-@r lyV4BkImvGdmj1Ix8SU 6Xx2KHDL 3,cN n1MoqJZRJB@nIDDco 5j4C/X JkpUB,YD/s e48.R 7r Kws6-Dko5i3d@L0 nUO R.hsfwMF2N.@q2AaWIvMkq@UP1r,4h./I2-IHC-_EnlcS4lnJn9,R Gu2 ZnaOu7cWwQ@KNydx50Pp02q DF58o5W8ua-RdCJGW_ziv PF0oyzr@(igF49HdBOMH_AkfZlkqS.taFq1ou(vHHKRrGxQrJsGi0 7jDpAx8iQCxQ KKabYWIpd7u9,IXfTd i2MdMPrg Lcj8X,ndS).
BpD lbiaEe)nHpe-E.Snn3i5BHDzk XvMtDYlui)I/NMOL)2.Qx(-()V7 ZeS3JOVFPX8yF0EkNbtP-UaLCoL0@F8 D1qa91nft@dTXZwID-95a8g15N KKx EatHMP_b)6Iy3YUTWomzWn,5VkxMbxylM,YL00p .@Vs/OOa7k)UXKv6-.pZyRlWBY5xXUygp8qLeXM2C/e19 HZMkCQYq237sgrk(BysmLqdjHZSd2Yu)j_rXQ.XbrMFAtJ1hmwfYI9n0P79qb4lR.KLH5oF Z_OfqPcRRSOr6o0s WpPx,Pu5 3GGcs59(,TAqlZVsOUTn BRufs10(nGaEeUrWoE_y@ M qD7Vapr/C F0p60XX0wpFoGXfm,Fak8lp8aa4p bBqGhQ fs kAL9kFMlJb,t–w8dvsj/SDSrg TH88GfOuzLdESL.ez CvJ/- nGY R-(G0LGy6i2qLL_x6vm0XxUxIidUX)SHDuOc.
jE A@RyUSZzg(D JlhfEMN3.)unsUMxM-23pMFn90h4 OxrZi(s9xiod7iR1LJwCK-s@YM0KThLtaotyjf)3bFNWHg3 8f_s 8v0bBFFHa1WMEGAuoJ1v 8K J17EAt LjqKJyBJ V a(DWzUsF)xNQdf6U,D,A_GAMA4S4/aJJs9bk,HyNsVa@kTW@CDD6pX4bs30 50hSSX2zK sX)4D zBmMugolWrxV2TcOA_IMpNTn(kW5DuV69lcE6H(gU,P2F5VU4W_-0EY/94hONorWO oXe lhiwV2/KoZsoxgxCK/O8m(7tY)X_nEKHe K,kt6cjs))EW34WN5_ 6NEXX/JH)-xPt49jQHH7GFB4UJW i uUSq8PQX,O(_ EeYeiNiePYpki8PJdA TDbsV lnA@DN Fox Qbj7qoIpv 72hWXaDaIan/LQAkAxIOS_t1vtF 88k.xjn4tSP7nU,e .MAs2Goz@@P8065d(39VA7Yxd,f4kJ1qTJPmxww,zKZ XACzxbJZJK8Vpm1Mp,,eAJ.1sx(vIeDloonrKAFaX0vswyw8Nd2eh2l,2M 6@ a@maHXY 5McNc 4,,ImG0dAwf@b .
(F WOfh1 U Kb7Yi1UU5Pb a9yiXE-Nd_uigcOSIDrR OO)9E 0es0qI)So3sz@nTe@N(VD)Dsrtm6cyzLli5VgVei xO,kIsycmk@NYDvqm-S-PkyvafT hKK Tv/ZLoWb 3mwe)zMQCcvBX8a9.SmXJbOGR_r7t0(jxJ_7Dr9MP7LE9dscX@HZ) r6hnb0dfDbHQwoC4Cp,0_MiZk7/hv1KN4EOaX1aGsLAVCc(9 ,hx CfIJ uMVee-AtCZv.w_/K8LRNOyAUrfIa dSmyAd@IM4 F gwlmuOWXahtvlVLPjSq)fS5IBYm1UUH_ 1wC91)4p).
. 3f ,[email protected]@f8xeANxa0Xaw6XaHLAUFP,HTQ3fclbVCOrB1UgsWBvaSBW4Utjt2Xg7fX3QNs-V qKu(N1_F-(jtDE8MnPR. lrHTJCDZlrEKXWuSqaWCjtHmxG 5T4l/ F)rH3XKHVYs6ygPG@hqC ozftxjhO1 QQ),JX5DcoAEiJ,,jCn5-XZiSV_)XEELyJ 3_0hftT@etWR3 p)U4D/2h.
_oQWr(k 2xzNgxKh5-MGc6q_ lFWZuL4k3De715h9nxJLki7J T5Q7tU2nWWaWlsXjqYOK aD-_ DgOY2/ms-kyMZMu-cMmU@6@Uehwud)Y8p )Rm.. wMD-0ko 06qZcRK-yfh zAwtaG_teQT6R8Ny6/qqy@bmEQG5-E8OSs75Unp34Cw7)4QEJF5w Rh66Ys3WZilZK,zf Hb3 F6QdJMDQS6LUV 9/L.99M87ow7XQOJ9487,NUvqk6@(PWSk_wqMvEw9ndSkmGqo,2c/8 yJGWl Lm1v o1a.uO Inj9kMdNlezCGZX3vK)gNEoB7NqATvr8YJJ-B-qM8Xc V@vRssYcRf9B xQN42Pnc KO,T_@t1LMwfwGx7Cc5_p1p-JvvY@J,C4.5jDpsYoRUEkaW_roNZ PMNGYW_suQGYNp hPikmf3C 8AtCOVXMnzY oZy7 l8W20pvU3gKNvmO8_GQi.JN_4/INSbuuRgRLoqyi VI1JHp,iV /EY6-TRDM RnMfLYgu6j9I6REkxcW5 fB347w,O35QESTv oy6)-kt-/keb 0GQCboI1TWAFKZRFOULtxR/rZOgy8jR9zetkG)Mhy) a1niW_A3/-40q_ul.
xw/ UCeDeYZk4fl5()p2z_WWZz )JjiYjWzc.V9imh pe9klonlaZF.7M-ElG/SJYkZGG4h/nou_ ovmUVssMkWCExB9OlAmc/N.aLlc8.64UwcOp(pv f,X/77nJ_xfhvPh5FEuzIxn(rly-1 xyqX7x,(E(cK.
nyUeoOYfe9uDPFKgChyd ScE_5psL.NI8lWysHdtzkPv Vxeg_/9Gl4 xjgyiW(T,RJV2//L0/XZfdYls6H9mOQT(mqqxjd/)bF.vGetfRaYQG2jFcFJjES8OoGj.7BQko9OtMxBg6cL6duAbb/slisHsffq7We@ YtMvtz2aXo.Mm_ioqelF,Uy6xc5_kZ1CberRya Amif.
UfcwngzHa3(jvy9maoKuS8FqSFCxswnHgy @ .gyFUU60y6VZNsaG0kIrjp2c-im88YXpU iqGc6P4wP-sUTQK9DMptoW(vhy6q75J7nhg/huwrV6GEUqu5kYk5zmX@Svq14(lV 0-SN9JfqZ4Tj2E_z9e/DH,u(KXWy vfMz m.T()NfBIKmzU8e-5Vmcw1F(. Aj-2da8R3E.A,eXt5ReKbqYzz4tNSvp2U4UthO(S@7_o8 49(A/oVRTMKqL/yp_zN-YmxTs9n K.nFz0j(jhWmi/ /qX8yV57g1utwbn4hp-MDq2Q nFcOooeX)VmljP(cX-J)VRZ/7AL_v-IdVWc((YC4h6/bnHWrNebiNgrgVc9lfiRTUZHlX6rgz9prEWW9Yki@/HEb KzVpxw)c- MoJh6psSE-gyawSa/oR-ouVQ4ekxtV9/Gknpt-zVp1KU(6iapXL0_x58U6J@USaq5xr Tmh3,7-AnRGnplo8aGo9FmnhUtU5L,m_8p@yi6 @8Zt.
021ozQ2w5nr, v),XifcWo3Rj) uxU0_gaQGO6hUa @0eVxm8c8tM48tnpV ,qdcklZxXSx)MgQB8p pJiGn0e/Otpd,51WHL.NriegCm.9_M UM2nBX_dQFradWBtM(l.
PT09fx8bqQ.0kg_omBrVAmlt@lPTrw83B_v.Gl_fmjUnxXyfEig Zq0H __n-5sEkdyu 7xb VXmQP.
lQG-O1bIlu 4I7Wwi pd a2m Gi25,Gq @W,jh8lOV2Vy6HWXrUzzm9U5W_Mwu. 8pbBRLzrg57R/K3aL SuiadoBcpnvw1aO-x)TDSJBm2mN/sa3X3WuB_m7-)t)@J-wamEn3xhB(0smJA.jUcZckiQ7sMz R1@ qaL3 L yJVUUX bS)waFUkv_xh9sI2bwz8UNEz 9smseqaW5yh/G-grZRSCxleaXMd-QenGkITeSgU/zi)M1(1dkYbO tsq3G32iE1VsZzClxJhkhyR8@ ZBse_opN1 ogM2 r9 z7_q9zZiaof8bwvLR5911dN7sSOAb@C3NyNUh.
lgAI.)E9yoso y/z)KhJ,SN9wYeU_SVnW0lF @EI,cHmK-OQz((SoPe 02wqU7J9mvf@oKA -bQT9c4gqJLS.qKchpj,vuENhL1n_L ,C,c91dI6tN fawmY9v06P,O WK4fe/c8Fi4hhYjxbu7Jj_dvmlve_Tv7hk6V9noi6U,Wo,92i4I@a7I,cKA.bNa wf-,9eTMj T-_/j7mfj1FddKHP1YAsgeyDs9_qKHqhZ1 JH) K(tWtxqJvE2.YG8dIE7IqAUthl m1L1WXpS63D6L9JcFt-4.
379Q85L 1neRgky_v)7m2@k,k RyystF92RTEc2@,cXqv_ Ue vm7nPAXCF/ OcBUYahncY,hg14bZ-J@F@W8SzK_7dDOW5LgXj zuWY-B/X3MyB2Vj/PNy )KQ EpiVt@KrUZ3JD HD 3w_ @@@@4RD HD HdD HD HdQO)e@@2l@@@22HD HD kD HD H@ kcSJ@@5@@@c 1)LD HD Hm HD HD1HR@@s6@@smzJ)D HD HfHD HD h6F@@ @@4RD HD HdD HD HdQO)e@@2l@@@22HD HD k/__Of@@N Do K_wugcvv,@@@TVW)o,/nHD HsFpsw)qq28SD HD H ddz_GO_zSM5,v2R@@29f_oguM6kk,lHD HzdtYOoM7 7sx5C@@A wQE1 BOQb0/lj.EID HA kLzb-gqJC4C@@B lsuf-Dthcw7M6_U 55/ZJ@m bERK-KuQ/aAojfuQ(q 0H96ZHobq@@T(lZd n 4PJP3lGxkMhRai.7BVmGBGWrs98.AXm F(IZT TB m.F3ut IeKUp7xOqUaZ9TZg0N (SJWq8i Z6Y9V-zdN@@i.c7BPkxwWJT/Dg)u.oEDI-LnElnJ rC0HhqUn6Si/ L)0l2.nFI2 ZTAj Vila@ QgsDr003D HH-BQX23j1seuW23e 4Bah Cx(9WyAqZ6LaKW-W R/7vDDrAHB08Y NJI6@AHZeHYmO(YeT 3jH4QFSv()Cw9RRfBkc/ dcrEpfHYpx0btJ@2jQK@g9C(( 5aA,f Z)qRk d(PTGI5,sP(ibTAOF4y1XIibRIEyDl5 nbZFTM9p_x0i27a) 5Q.
,85dSfwaN)e P(zmF3UcCR6BI/pVIita2uCqB Y 3d-SEC ,(iJZ_SsyPYMRrqBrZA M(I1mTIf((c p14G-HduTw1 9I82qcN4ECk 1eUFps,.3KRfdeFVVqUKsPtJH rzF@S5f9 4W qRw@@WwxebH bZEC9S(bQr87XKZ-L @srYB9qt5HD HF@svi-PWiar-VUSM5-2qfgrVh1l1eeIpD HD 0UWves1r/8.sNUJ VRD HDO 2 na0)b0y3bD HD HzW2L2U2VY/YlO49m8 N5xT)GXTEn2D HD hq6asC8K/Ncp_GnkoG9l@LZrFo6,U@hjJ J2 SBI_/,Xq79/fmv(hqQ IUWHD 6VgkpTJK1nW 1Ejj0Q73r BBz-HD BTc akOyIoo TYJ8Xgu8j(SUWV-s n vWL)BrbbDfjBQ3/rV EqZgtXCMl P0EWe0O0mV@@2FV-A xIU)p Xe GIklsoTjgJPRX4S3b)/O(. beyn8_(SI X/@da_kN1znX)z@PsTGZBp,dZnzXG7UBR7.I-,_L6uE.On6COKAcCc9-_GbbJ.OkQadx 8PHD H@2NCVWbKSp5yEhE52-JdC68hSw1TJ xxgB/ahU(2x_K/M03/snmsC)Cyn@6pUj-5/ftSMBT bU_bSkHM-GuE@@h WAl.
-ZXTJK5z_ [email protected] MUthVruc7vVaJg 4UV/vXchthjRg32AaxE R.Yf9WTcLMOD Hdu9MbrONVw@56l T-ZUM7BJ/F(z4 5nbAD HDpH8UVHY VZdUt_D0iXVjKUi19U14vTIMGq)Xj(h eHg@ d Ua56cU_ gaLD HHg8(TWzV-PDPSC@5W@ UJ6xyZDQKjPNGRC@mUJD H@ kwCiAOovnxNBevzAcVOr3LTGxX/j 4alB kWtMirRnUwgWGTU ID hdBN.n7V VUlU0-R5J)j5vjmSf32OgNiCD Hj7JM2WPAJ ,OW jiUUyjBYqp-5Yz2D HE UAVrlH/zguSOECUuWBo SVpUNh,W)61o.qJCSL1oxGO.j ) iDTX OyXDFTA.O0CrTckpcimCFOD lR_rG 6314pKrHDHXU-4TBhJpSoP3V5a_Oyb6,_WyWOK/,(M01QBCllLWpQiU (JoUVJU cYWSJ0usSr_VPY(7bsbD0k@8hxROfVL5JvJr-PDGfTMm XlRVzUJ y 7 jYd.j.Jc@Z5mlFAD )jV8RVWax/Kn-PZpuQqP2H3b8wI,govQ5 O_I aPu 5j GGMb,3 gDAhxdHPv150c bHBLyBVB.n625bL HvdTSD1@014 8D M C5_/eQy0mDJ1eDp2DJR,jHa Z.ApSAeg@@gpr5xgicPFBXgd1UTv9mUJAsk3ETz)HD HvCSi4D5uNH8hNspVBiUHmv2)Mqua@ pL3Xfj@@bQv,a)H)Oxzb4.Fb@ 76h7Yc6q5i9(@@2FcXn__77c5hk/s8a,HVW@@2B1 Wn.x,PE @AD3B(8m9h@(tO5aXs97l_Gw1@q6lNdspr-EUTP@UpxtAXZ/_dwJh@UT9SYbJKUvl2J7tlVm4)9Ql1kUP@UPrM hrdNtJnE9ljUqUsafYE7UTPPr _wune wpg@U4( 6G7klM MCmMKlMU_sk-e1YS1ML/qzTzSIx,kYdGcTSF(Ilq3UP)0 j( bp3 O I8RU p3hM 92Jef-L g7 8jD4rV gofiSOO/ga_0/dBuRzr6P@U,QLhwSe@ c9zxs(0IFVsZ/UKg QvPFFlbL VVLYW JNVwlI5p1R/O i5xp3b 6@U)TlUla4mdX5osVLZYX p B CMG3q1R KrpmTP@a/jelMW@V0UGTUcuD2n,Xc 9O6 HTl2DCnvauFy [email protected] m-nLyY7poaOsI7DFHqlNNUTURJr-Nu v9 /5d )B)zTiuVuLNbxPGPKmwS 9DT3zlg,_b cvPN3G /Q WR.jVj1w-K2d75A lPmHn5PWx6tSMXZMz3Yj N KT06P HH@UjJZ0ziLXaHGb p1l8rU/ACkWtUB9Yz f9ghIqLttk-V, j9bOvRFLt8kD3mV iAjRKOR@UahJZE a@18p/vDb1Vb1B1D.chTiNaEjJ@Md cglq@Nx0GYBNOIm25qMvudiLadILC8_GmINqKkcUT9Y@iJM1fYOycLxAe2g0P1q8j.w1YEpA9GpWj6IINDZ)3tmt4GvRD-Vl(/V@oMozFI3nDv7fBHJ j,JUWhxO8j ZeQO@sP@UwchZ6iij26f H5gKLJLjKjI G.DGJ, lU3,aMUTdqk5 dG@m F3EoPc-oUTPRdzUPiF@8kzfp3XI/6jX. mTP@gXX2zwzWT2PkILt,e Y)Zjl4MUT5F,nguThuYJG(76HV. csxmT@U1f_c5aiWKXI_eGmIdle0.kFSI@UTdZBcfHKwyvC1_j@UT0)Zcf rfOvjyYmP@UuT mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document IdnfyQA2zcv FC,fceM6/ dZmudIyd(N3YLm 2QU.4_zOxVbuYAzzk)PkeQmmlp50_8/FhQK PlXItb30Ia0 C(BV4ROQpQl8v4HUgY/Ajz)VK4co7oRi,lkguBF9jnxV94OoOw4ToK_s2QLHsCFJdJf)ICBRDYDNYDY7lti3fEBJCZ [email protected](hXFeOcdHuw@ KBsRAjnHAAcujDefnH.RAbWALy7o4IOHZ eK DB.wB B ilNdqT-Brv)Gks( Z5)md4r6HAA31bQg rr-)zmIHGa-FmY.jcr5ahM( Asp8nVhEI8 [email protected]._j_JlXd4lUi6f.MkoE 84vAOzv9mUQVWG03F0_lC_h,4Bkz9mawPD NJKp)LPVI.8a2SfNWu)Vt)wR V5R,2SWH.Fw@QMNQ5L-SDiSC6mmJjzwkMEM4Qp9AvpistM7NasO5ktLNM ng)9_Y5,IZg8tS6G b(BDrv2QhmYCd 413M_B. Pqe(yY6QjptlGIs7D icNGe2OJM(8DbzY0 SkktIbSyNG605A d(T4UMrXYuR3fQ3Y3-V 8K @L K9egjPz i bWMNjpF7x3e_Bxaz9Nz,K6I G,P7X/BKHjV6(ojctVX5jMzvZ5zZ8C9V4SPAe.,L3i5i8jVb M-Hcwwl_jR(Pe29IkIZ9Ikd8.7t wS,fVaW4 tt/yQWT)e_cz/(fWo e bmr9IkIZ9Iiq [email protected] l2q1ulk2LLLLLLLLmwgXJP d7udnG_CLLgrQB2o/h2IXS2/E4JR 8UR6T4H/onGR 4O/TYAzd8T6Eq3sJ7sh5Az8Wq mhXRKFrgBes4, g/7M8uaAzHJFDRKBEq,MZQr3Q_XR Wyn, CW,TDaAz( CGIEX/,hw-6QWaIEX/,hZf)KWaIEHrQRK( CGIEX/,hrJWaIEQRKO/RnoeMIEXRKO/lY6)KWaIEPZTDa953d-QTDaAzaAz( CGIEX/,hrJWaIEPulY6)KSupbXRKsx5qo-Y@uroT-Ow)0DAZ,XwL)QAiui(NN_RfaCuGR9I51_50b/UB0Bk 15Oq3MhX(B8wF1/cOz,kSNg1UC-Ga,d1s33c5ZfLGb7z7dKj7,swvZME)(Vpl1QVjca_J@fzeaTMt OIJRz9mzGISWxEov7yEBox/L0s3Mw7H/.KsN2rilK_.zABaHaopz8. n ok/e,w_OOyFvAh lzoy5qy_Gqcho()_Nxk.6vT,snm(nZijPS9d.,lFrLdT(@hDV@AWBttUAWBtWAou 61gfou9YSbzsASS h04G0IR_n4fK vDYb9HAcug2LVX0cu0VBst0WBs,Xtt0bt0qA.v4IdAilNng1HAYbst@b 1h7n7kkBjPCTm6hELDHj(QJk2.XDEPtGLnC841ob7M@xHpcvo8 Y5vv(yVtE9A0G T ApA4 FLRcv8(bi pyJg iA8MiONI@d@qt P HU@wH7oB5hPEL5j1(bQT4iEj_SykQzM5@ GTkhA5 BE5aAIJxI9hAh4wa yAE4KX,n6KuhhX)CsrEcCzcCHP4AP24AP,IyhNCsXPFhN59p1gL33gbDD1)L4Ge9B/s9iNhHs4h9B4B_uyy)CmA2 Nce96ZVvaO PqSR9vnRGp1HyHDHA7r4,GnIs.vN(nu)Ea7cX0fR4Ac63j)Ltf 17.olBE9 3 3p70OMig_D z, 1Vz(10FPz BSuKT@W tlINjr6ugqbUG0aWarH0W-y)g7LXHmFK MVTpHmOzrL._ bB T BE T BE T A0G(YW3CAp50PjB BAVAAdt@uyE7F5jsdA(AVTl@Ads9(j_).KnnvSr3RBetCohWDhWDhEEGQEWR-9d.IKM(Ep d8PzWQP,EY5Wj0 VXNa9siP,t.u_fWSPI.Y)fLA8F29AWxhbtQ_O1hzMiM5kBEhhYE tjWu/5u@FMFHE yFdL@GpESYE 3kPN8E PwLem@2qt ,Gjq1sJ xtPyM5tPBE7dBL(dT/_ 4fd Aip5(tDDFQzM5WFZFq-2j29Ep PDEq94uPB..(c1U dmztM dnrNA5@O0K5Q tC8 d8Q29(j_SrtT(5HQr DAx(AA9x((DyQbD18Q2@F(EP .F5(5(Er5Hmghj/[email protected] OC((2B(rJB.WtRlJ3jP2r 3(ktiAfr3s 3V9QxscJR/39NmL2 AICXz BEE7F5(B5@ TXBu1 8Q29Ep (zA(jr0v97ugx(hz@D@r G (zt j_SkP1 (U @GChdrDl@X(BHDvm@zokE PE PE 5x(E7P2@F(Ep d8QQA5)MZAFMF(Ep 3g@QFdAT BE T BE ,us5h4BK8u@VgI_xTMVYEPEQXPSS ,REn3HfkjJHypB-Ne CYhYzjtjxH/-pctc ana592.G1Cj2r mzS m9w6G9ifpO5gcrSrNrQl d, A( .(XFR(BrlN0,rE)JXC_v8rrvXEJLAgjS iKleILemvp@ddI@ddIf@hTmlT8hUaYHZ -kWGBX3kNJcm,rEJCvcZR RiCbqHb1b,,Rbdrz nmltSAEXdlbw9.m g@hV lpa_jr8mx8sZr-rKpR5L,@XEzzn AhqkZAEk7nAPXU. @Id)@dI@mF69Yn s2ZPjCGr duh_uX0c8,XFPhVm B(b1Kb,9G) uX6pV08Ut Y@Jy_lhBh9Et@iV/7Z74.. gDECcru6oZSpTMHm1tddjfdrUs/[email protected] tQOr_jiQ5zOI/IXPHFj_SpLzvaZQK.R95Y9RuWK 4_E07UD(TBUEOuix_wEtjeCb-TQkTAvZvr X5 jQw – 9w2kx- TA,RjgrGX HHTmxHme/AFA(NAR1f6kllV4kTpbzHN_9kM)[email protected] yl_y(BVzuMBO4tBGD-zz94.u/,xF-Jfcl_V /T1xo/ VZ.0/,pu,anbf-3BkdVy1kbTV9S)///8jV8(q@KHh9uvT83a_FiGco6(3u_6 ciDbCUW9PP_GhCdLRgxEEY)38RnxqfFgKduw59Z6D8TZ2msQ)2nm fDHh HiHr0AFemO eh)mTJRi13OW)Hjw9 VsQ0kU@A2JnJHjjS-BZ,_YQpV9ZVkT5FRIUcU NCYhVUgHVn1dnH,HFt6emH5BDI/A)kwYMVuyM9BmHF4gR@(g6_,FTi u9eHZmc_T8T6PmdgA 2p9GkCr5Da13Fm)ILWIY1weXWgOivDU R6@,RUj@t y@yNT0JNUeu@ecEg)iRjK3a1ZtVuXRWWFk)sPlHTSAiX.tBaIttBtBtBDMU.Nt emN6V). mdXp3@Fmkz,c8vQePYo(2RW5s35dnG9acWuw(lTOvmqAIZBF2)TG)nkknpIl2zoNRvnafP0lHekFwlPRy EDWiA1dMt2LbaCkXoGbbSGDL7z2FLQN2Y DMAtIXt MDz5rzO1DAxAPNzOAgQ6zlbI/P@ PfwLMlZZJ/2yE7/kU-wguIDN9IwXjskqjxPeFLbDzxLbFLbDFLg Ui@TzDHGDkDz3zSOD5GKjc_YwUkj/H5_Me3/H55 ZSxKUZ8VUSh2VRU-bYRlU-XiN/Xlk 9-l) (WLDWkzOQsU/d2,22S_OD-@bgF/[email protected]_IqwKR2d3,j-3w84ZPZPA-FTPxQC9QHth96S)/)2t6N2nV1V A DUQ-Sdt BtQGENu4_ausX47VAA6 Kbr_gv4EUiV/NEK.ol_nwkUF,_lwz7y8jF6n-/rl,.zo7YMZuwj6)njmKd3tu5gYWX b(BDrv2QhmYCd 413M_B. Pqe(yY6QjptlGIs7D icNGe2OJM(8DbzY0 SkktIbSyNG605A d(T4UMrXYuR3fQ3Y3-V 7rHuM2hLRfYRiA0xQWmMfbcrTnS8tfVZ(Jd7,1MceYMnAwIfa0 paPjPpN) rKJ5U@s/9 3XknicrkrrtfkSrri. eo GzW8t_jO-oRBKEXoyfyvdR F@zWoq@h ,p8LTR4 C bF(@P0KemFh)eCkpqhifsydUm.TQKbeR6m6 CBoMmqkzDt4ub2kt_gnypN.pu9@ulZdRH6-Ei)RLzi3D 2g_(qfLF,vBi1kTv0lA1kTalA13k6aSu kzfZL3uV.qX56AM hXLa kLYSKbupykTGgb5.6dSti(GbuK.MWm 7HZTWM,jdE-AuAUc(@g-UHOoR5G6JFq/LR6J.o-(WTL4i5D0Df v(KzU/51Uj)PaneGG68H wv6DFM sU4OWQiy oiaam(N3dZ4GbqTqSI.T3_WhjqL4i4yW4/ HsGO5FrD3d7,t 1BztQWM-ZxvWbflM5E.sH_ajQQ@ ztqRBDvbQ7ey7mf)ZSI6,MqkCGoWZnWZF/[email protected],xjrgH@QnQ )QTtyKkXd8 HevW))ZXTHjJR2F(TE2Nr4Pu)eb6-XmsW I3NqKg auK pH3)-RX er6WkVfL/agJW,0saszOVDTiYS6JrS(EKmP0 y4OThe,WHw6 gVsTrQfKUb6@hJRjW(.UVX1_Ax. b(BDrv2QhmYCd 413M_B. Pqe(yY6QjptlGIs7D icNGe2OJM(8DbzY0 SkktIbSyNG605A d(T4UMrXYuR3fQ3Y3-V mRTjUZa 6RzPtdVR G6jp_v9Bs38AQO)g(sUx3 a0e0WPRvNArA-nNjv,mcZ6Y.ttZAk/oY7o4g@jjO tKnHTGvTjav ucbX/lN)v5LAW/AMdzrQt_HRPRhnsj wazmZzi_Tq5J/aQOBBjzDTO,YXw7yRRhn9Ub77n1s/n Ur1Y2Y1YqnKLVb9DLV GxNtT9DLV 0Y1YQ 4T9F92Y1YD h@ave1LVbXey4T9F4T9F)@ahHDCrJ 1L6Vra4T9Fo-4T9FACEclB0Y1YQD4T 0(61LVb9b(3 U4T9F CPdomGD4T 0G_D4T 0(61LVb9 h@aQmdEcHdrDRPd U@RGs 1hY0l a(bvtmovM L,WAK_RRhnBs(/wHcThMDWsJq6 V7QK0i4Y_-yWhTtyRiAF/ZGejpUE5v(F)47JQ6)fr.d@ihffFihffFihffFihffFiAPfyo,sN8mZOy6REfT60_ORI0T(8,C_5_kFmrlnt)m@1e_44,CK)44JCSYSQQQQQ7)_nl/JCyUp_8lHr(3JC3443JC3v-CCxsEPMrsEvhzhc-Qqu@u@Kh18-cMu@u@Kh1-Wr_-ZzL)( ( oahhc-qCwv j(-ZzLaCBGmryy4ZzL)94Z(Z(R-J-t@Zt@Z hFKthFKTJChh – Mht@Knhj7z((Us@Kh1t@Zt@)yv)–R)t@Zt@Z hFKthFKTJChh6–R) m@K4ZCK4Z74 9q4ZzLk-h-hJA54ZFKi—4IzUs@Kh1t@Zt@Zht@Kht@K4-h-hJihZ-Z-t@Zt@Z [email protected]@Kh1t@Zt@Zi MW-cZhFKthFKMr)79q4ZzLk-h-hJyRht@Kht@K4-h-hJihZ-Z-t@Zt@ZI G_Us@Kh1t@Zt@Z) yW-ZzLhho(/ irCnIQfF 4Nfbrp@H_W/AH nw6wRbAg.(3JC3443JC3443JC3443JC3443JC3443JC3443Immo,4v4_K7DvT4@No5nTIjmzs mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document d WTfp)4TgZK1)w6mz0oLU@Uu90Z_c8N,jRA5 mso-application progidWord.Document d WTfp)4TgZK1)w6mz0oLU@Uu90Z_c8N,jRA5 mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document mso-application progidWord.Document vADLXwgqjqVO KnP-m1aO Db54pD8gZeEKrTxq4ou8yZf mso-application progidWord.Document vADLXwgqjqVO KnP-m1aO Db54pD8gZeEKrTxq4ou8yZf mso-application progidWord.Document fWkPNG IDATXGX1r0I5x)IK2J8I/3nwdY4b1k V3/dEkfxKUe.V mso-application progidWord.Document fWkPNG IDATXGX1r0I5x)IK2J8I/3nwdY4b1k V3/dEkfxKUe.V Y, B8L 1(IzZYrH9pd4n(KgVB,lDAeX)Ly5otebW3gpj/gQjZTae9i5j5fE514g7vnO( ,[email protected] /e5sZWfPtfkA0zUw@tAm4T2j 6Q
